The Covid-19 Pandemic: A Public Choice View 3031249666, 9783031249662

Table of contents :
Preface
Contents
Chapter 1: Introduction
1.1 Public Interest Explanations vs. Public Choice Explanations
References
Chapter 2: Viral Mitigation: Weak Theoretical Underpinnings
2.1 SARS-CoV-2’s IFR Speaks Against the Equal Vulnerability Thesis
2.2 There Is More to Immunity Than Antibodies: The Case of Pre-existing Immunity
2.3 Is There Such a Thing as Asymptomatic Transmission?
2.4 Wishful Thinking Part 1: Zero-Covid Through Lockdowns
2.5 Wishful Thinking Part 2: Zero-Covid Through Mass Vaccination
2.6 The Fundamental Problems of Lockdown Mechanism
2.7 Lockdown Failure
2.8 Pre-existing Immunity vs. Shutdowns
2.9 Modeling Drawbacks: Poor Inputs, Poor Outputs
2.10 Displacement Effect, Lockdowns, and Focused Protection
References
Chapter 3: The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria
3.1 Laying Out Mass Hysteria
3.2 Mass Hysteria Intensified and Drew Out
3.3 Mass Hysteria and Social Desirability Bias: Panic Institutionalized
References
Chapter 4: Tradeoffs and Knock-On Effects
4.1 The Wrong Dilemma
4.2 Estimating Tradeoffs: Established Knowledge as Guiding Principle
4.3 School Closures
4.4 Economic Devastation Part 1: Deep Recession
4.5 Economic Devastation Part 2: Increased Spending and Inflationary Tolls
4.6 Isolation
4.7 Unreasonably (?) High Excess Deaths
References
Chapter 5: Public Choice Theory: An Explanation of the Pandemic Policy Responses
5.1 Can Voters Ever Be Public-Interested Agents?
5.2 Voters’ Ideal Point: A Conglomerate of Expected Utilities
5.3 Politicians as Vote Maximizers
5.4 Public Choice Theory Applied to Outliers
5.5 Not Ill-Informed Politicians: Public Choice Theory and the Precautionary Principle
5.6 Public Choice Theory and Mass Vaccination
5.7 Bootleggers and Baptists
5.8 Pandemic Responses’ Popularity: Weak and Hard Data
5.9 Budget Maximizing Mechanism
References
Chapter 6: Epilogue
6.1 The Sound Scientific Grounds of Public Choice Theory
Reference

Citation preview

Studies in Public Choice

Panagiotis Karadimas

The Covid-19 Pandemic A Public Choice View

Studies in Public Choice Volume 42

Series Editor Randall G. Holcombe Department of Economics Florida State University Tallahassee, FL, USA

The Studies in Public Choice series is dedicated to publishing scholarship in the field of public choice and constitutional political economy. The series includes research monographs, edited volumes, textbooks, and reference works in all areas of public choice and constitutional political economy. Theoretical models of political processes, empirical studies, and case studies of political processes and events fall within the scope of the series, as do volumes analyzing the impacts of political decision-making on public policy. Public choice has been well-recognized as an interdisciplinary area of academic interest, but public choice analysis often has been absent in public policy studies. Applications of public choice to subfields such as macroeconomic policy, health policy, income security policy, and other public policy areas are welcome. The target audience of the series is broad, ranging from academics to policy practitioners. Projects submitted to the series undergo evaluation from the series editor at the proposal and manuscript submission stages. Additional rounds of peer review may be required at the series editor’s discretion.

Panagiotis Karadimas

The Covid-19 Pandemic A Public Choice View

Panagiotis Karadimas Athens, Greece

ISSN 0924-4700     ISSN 2731-5258 (electronic) Studies in Public Choice ISBN 978-3-031-24966-2    ISBN 978-3-031-24967-9 (eBook) https://doi.org/10.1007/978-3-031-24967-9 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To my parents and my brother and To my regular interlocutors, the incredible people in Korai square in downtown Athens

Preface

My initial intuition back in March 2020 was that by viewing through rose-tinted spectacles the pandemic policy responses that were first then implemented could result in being grossly misguided. As public choice theorists have long argued, politicians are not exactly the benevolent individuals that care for the common good and have as their very first priority the public interest, as the public-interest theory suggests. Instead, they simply act as everybody else, i.e., in order to increase their utility. Given these propositions, seeing politicians implementing policies that were unthinkable until 2019, my intuition was that there must be more than meets the eye, and the panicked population gave an important clue of was going on: The sweeping mitigation measures came in response to voters’ demand for action. Possible inaction or even targeted action could have been conceived by the public as ineptitude, and politicians would have ended up worse off. The public was in a state of mass hysteria and failed to rationally assess the dangers of the disease and the dangers induced by the mitigation measures (especially by the lockdowns) and so conceived the policy responses as life-saving. Thus politicians and bureaucrats maximized utility at the expense of the majority of the population who ended up worse off due to the collateral damage inflicted by the measures. This is the central thesis developed in the book. However, to make a convincing case in favor of the application of public choice theory, one needs to compare its predictions with the ones made by public interest theory according to which politicians followed science and acted in favor of the so-called “common good.” The book, therefore, is confronted with the task of testing against the empirical record which of these two theories best explains the pandemic policy responses. The conclusion is that public-interest explanation is rejected, while public choice theory explains very well the policy responses. Politicians acted throughout the pandemic not as public-interested agents but as self-interested ones. Part of the argument presented here appeared in my paper “Covid-19, Public Policy and Public Choice Theory” in the fall 2022 issue of the “Independent Review”. I thus thank the Independent Institute for granting me permission to reuse material from that paper and especially Figs. 5.1 and 5.2 that here appear in Chap. 5. I am also grateful to Chrysostomos Mantzavinos for many helpful discussions vii

viii

Preface

and to Bryan Caplan for illustrations on the applications of his theory of rational irrationality. Last but not least, I thank Lorraine Klimowitch, the publishing editor at Springer, for supporting the project and for the systematic communication we had throughout the publication process. Athens, Greece

Panagiotis Karadimas

Contents

1

Introduction����������������������������������������������������������������������������������������������    1 1.1 Public Interest Explanations vs. Public Choice Explanations����������    1 References��������������������������������������������������������������������������������������������������    6

2

Viral Mitigation: Weak Theoretical Underpinnings ����������������������������    9 2.1 SARS-CoV-2’s IFR Speaks Against the Equal Vulnerability Thesis��������������������������������������������������������������������������    9 2.2 There Is More to Immunity Than Antibodies: The Case of Pre-existing Immunity��������������������������������������������������   13 2.3 Is There Such a Thing as Asymptomatic Transmission?������������������   16 2.4 Wishful Thinking Part 1: Zero-Covid Through Lockdowns������������   18 2.5 Wishful Thinking Part 2: Zero-Covid Through Mass Vaccination������������������������������������������������������������������������������   21 2.6 The Fundamental Problems of Lockdown Mechanism��������������������   27 2.7 Lockdown Failure ����������������������������������������������������������������������������   30 2.8 Pre-existing Immunity vs. Shutdowns����������������������������������������������   34 2.9 Modeling Drawbacks: Poor Inputs, Poor Outputs����������������������������   35 2.10 Displacement Effect, Lockdowns, and Focused Protection��������������   44 References��������������������������������������������������������������������������������������������������   48

3

The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria��������������������������������������������������������������������   59 3.1 Laying Out Mass Hysteria����������������������������������������������������������������   59 3.2 Mass Hysteria Intensified and Drew Out������������������������������������������   64 3.3 Mass Hysteria and Social Desirability Bias: Panic Institutionalized����������������������������������������������������������������������   67 References��������������������������������������������������������������������������������������������������   69

4

 Tradeoffs and Knock-On Effects������������������������������������������������������������   71 4.1 The Wrong Dilemma������������������������������������������������������������������������   71 4.2 Estimating Tradeoffs: Established Knowledge as Guiding Principle��������������������������������������������������������������������������   75 ix

x

Contents

4.3 School Closures��������������������������������������������������������������������������������   78 4.4 Economic Devastation Part 1: Deep Recession��������������������������������   81 4.5 Economic Devastation Part 2: Increased Spending and Inflationary Tolls������������������������������������������������������������������������   83 4.6 Isolation��������������������������������������������������������������������������������������������   88 4.7 Unreasonably (?) High Excess Deaths����������������������������������������������   89 References��������������������������������������������������������������������������������������������������   91 5

Public Choice Theory: An Explanation of the Pandemic Policy Responses��������������������������������������������������������������������������������������   97 5.1 Can Voters Ever Be Public-Interested Agents? ��������������������������������   97 5.2 Voters’ Ideal Point: A Conglomerate of Expected Utilities��������������  101 5.3 Politicians as Vote Maximizers ��������������������������������������������������������  104 5.4 Public Choice Theory Applied to Outliers����������������������������������������  107 5.5 Not Ill-Informed Politicians: Public Choice Theory and the Precautionary Principle��������������������������������������������  108 5.6 Public Choice Theory and Mass Vaccination������������������������������������  111 5.7 Bootleggers and Baptists������������������������������������������������������������������  119 5.8 Pandemic Responses’ Popularity: Weak and Hard Data������������������  122 5.9 Budget Maximizing Mechanism������������������������������������������������������  123 References��������������������������������������������������������������������������������������������������  128

6

Epilogue����������������������������������������������������������������������������������������������������  133 6.1 The Sound Scientific Grounds of Public Choice Theory������������������  133 Reference ��������������������������������������������������������������������������������������������������  134

Chapter 1

Introduction

It was constantly claimed that politicians followed science during the Covid-19 pandemic. This suggests that they acted as public-interested individuals who took expert advice to protect public health and to promote the so-called “common good.” Another possible explanation of the pandemic decision-making is that they ignored science and simply acted in order to maximize their personal utility. This is in line with public choice theory which holds that politicians act as self-interested persons. On closer inspection, it is showed that the latter is the case and that the pandemic policy responses were at odds with science, served the interests of politicians and bureaucrats who were at the helm during that decision-making process while inflicting serious damage to the society by and large.

1.1 Public Interest Explanations vs. Public Choice Explanations Different theories compete to explain decision-making and political regulation, and each of these theories has its own postulates which lead to different, mutually exclusive, predictions. It is, mainly, public interest theory against public choice theory. The debate is old, albeit it constantly appears anew and understandably so, because politicians’ actions cover a wide range of issues, and explaining each time whether they acted along the lines of public interest theory or in accordance with public choice theory requires repeated testing of the hypotheses these theories provide in each domain of public policy. Public-interest theory is considered to be part of welfare economics and its purpose is to establish that government regulation increases social welfare, or at least that it attempts to do so. It thus rests on the assumption that governments act in favor of the common good and thus regulation takes into account all the relevant tradeoffs and seeks to promote the well-being of the society by and large. More precisely, it © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Karadimas, The Covid-19 Pandemic, Studies in Public Choice 42, https://doi.org/10.1007/978-3-031-24967-9_1

1

2

1 Introduction

assumes that government tries to achieve the well-being of the society by correcting via regulation the failures of the market to do so (Hertog, 2012). This cuts both ways: It regulates the markets in order to protect the common good against possible adverse consequences of the activity of private markets (or of individuals), and it also regulates in favor of some private businesses in order to protect them against competitors and consumers (Mitnick, 1980). The government is said to remain neutral with respect to private interests and regulation aims at improving outcomes for the society (Pigou, 1932; Stigler, 1971; Posner, 1974). To achieve the optimal policy, advocates of public interest theory claim that governments are advised not only by their fellows in the political domain but also by experts from the Academia so that robust knowledge will be at the disposal of decision-makers and the policies will be even more likely to be beneficial to the society (Jones, 1988). Public choice theory rejects nearly all these projections. Among its founders are considered Kenneth Arrow (1963), James Buchanan and Gordon Tullock (1962), Anthony Downs (1957), and William Niskanen (1971). The main premise of public choice theory is methodological individualism and suggests that individuals act as self-interested individuals, and they engage in utility maximizing during their everyday lives. Politicians are individuals too and so, as public choice theory proposes, they act as utility maximizers and not as public servants that strive to promote the public interest. Individuals who work for the government, as are the bureaucrats, also engage in utility maximizing and thus they are not public interested agents either. Thus, governments are not groups of people who pursue the common good by straining every nerve to find the policy that increases the well-being of the society, but bunches of individuals each one of whom pursues their own interest (Shughart & Razzolini, 2001). Thus, the theory of utility maximization, which is the standard economic apparatus, is applied to explain the actions of individuals in non-­ market contexts, as for example in the political domain. Therefore, the main predictions of public interest theory can be succinctly described as being of the following sort: –– Policy A serves the public interest (government advisors suggest so). –– The government implements A to promote prosperity. –– Eventually A makes the majority better off. Public choice theory predicts the following: –– The policies enforced are usually irrespective of the policies that serve the public interest. –– The government implements policies intended to make the members of the government better off. –– Eventually the decision-makers (politicians and bureaucrats) are better off and the majority is usually worse off. Testing these theories against the empirical record requires estimating whether the declared goals of each policy are attained and if so whether this comes at the expense of other groups in the society. Since all actions come at some cost (even negligible), the purpose is not to examine whether there is a cost, for this is nearly

1.1  Public Interest Explanations vs. Public Choice Explanations

3

inexorable, but to examine whether the benefits are well balanced against the inevitable costs so that the policy that is implemented can be said to increase wellness in a society. This includes an exploration of the available evidence vis-à-vis the effectiveness of governmental interventions in each case as well as taking into account the available theoretical knowledge so that the evidence in question will be interpreted, where possible, in the light of this knowledge. Since government interventions include issues related to diverse fields, a careful examination of data and theories from vastly different branches may be required in order to be able to asses which theory best explains decision-making in each case. Note further that rejection of the one theory in a certain domain does not indicate on its own verification of the other one in that framework. There is further work to be done to precisely illustrate why the other explains the decision-making in question. There is moreover an important ethical aspect that is integrated when it comes to policy making, but in this work, I try to remain neutral with respect to value judgments (at least to the extent that this is feasible). That is, a policy can be effective but unethical, and on the contrary, it is possible to implement policies which can be supported by ethical arguments but which are at the same time disastrous for the majority of the population. While I do not deny that value judgments can determine or co-determine effectiveness in some cases, it is I think incontrovertible that many a time the effectiveness of a policy and the motivations that guide decision-making can be evaluated without appealing to value judgments. For example, when governments regulate the minimum wage and raise it (which is what they typically do), the standard economic theory highlights the tradeoff: other things being equal, unemployment is likely to increase, but at the same time, a chunk of workers will earn more money. Since we know that increased unemployment rates do not increase the well-being of a society, we know that regulating minimum wage is typically not a cost-effective policy. Since it is not a cost-effective policy, it rejects the predictions of public interest theory and we may look for possible applications of public choice theory. But to evaluate whether this is an ethical decision-making or not requires value judgments and one can make a moral case both for and against regulating minimum wage which can be irrespective of the evidence-induced assessment of the effectiveness of the policy. As for our task at hand in this book, therefore, moral arguments are set aside to a great extent and what is taken into account are the predictions of each theory and how each theory performs against the available theoretical and empirical knowledge. The policy responses during the Covid-19 pandemic have been far-reaching. Governments across the globe acted in unison and enforced nearly the same policies with approximately the same intensity. Governments and their advocates emphatically said that this was the one and only way to save people from the onslaught of the novel coronavirus and that even considering any alternative poses an unprecedented health risk to the entire population. In support of their claims, governments argued that it is the established scientific consensus that backs up these policies. Governments reassured the public that sound scientific findings guarantee that Sars-­ Cov-­2 poses such a great threat similar to which humanity hardly ever encountered in the past and that it was scientifically beyond any doubt that lockdowns (with its

4

1 Introduction

components), mask mandates and mass vaccination will stop the virus and save many lives. The possible side effects of these interventions and primarily of lockdowns were, according to proponents of this view, negligible, for the risk of contracting the virus is so great that wipes out any other concerns. This begs the question: Is this an example of public interest decision-making or not? And if it is not, then why did these policies put in place? Did they serve the interests of those who took the decision to implement them without considering the detrimental effect that these policies could have on many individuals? The public-interest approach to pandemic policies involves the following predictions: (I) The risk of catching the virus and dying from it is so great that all age-groups are threatened. (II) Even if one does not die, the woes of the long-covid will be so great that will take a toll on the patient’s health in the long run. (III) According to scientific evidence, such a deadly and dangerous virus can be mitigated (if not eradicated) through lockdowns, masks, and mass vaccination. (IV) There may be some economic downturn due to lockdowns but the economic outcomes will be the same regardless of whether governments implement lockdowns or not. (V) Based on I–IV, decisive governmental intervention is required. (VI) Eventually, the majority will be better off. Chapter 2 undertakes the task of testing most of these claims. Predictions I–III are discussed and they are all rejected by the available evidence, and it is moreover shown that there was no theoretical grounding to support them in the first place. Challenging sloppy applications of science is also part of this chapter’s agenda. It is a common pattern among modelers to posit assumptions with frivolous empirical grounding, take them as incontrovertible truths, and then apply them to various contexts to see how they perform. The methodological drawbacks of these studies are highlighted and the results to which they lead are questioned. The central conclusions of Chap. 2 cut against all the medicine-related public-interest theorizing: first, it is shown that lockdowns are either irrelevant to the viral trajectory and that there is strong possibility that they increased the virus-related death rates. Second, it is argued that masks also do not impact on the transmission rates. Papers arguing for the opposite are based on the sort of modeling just described and lack empirical basis. Third, mass vaccination is unable to eradicate the disease and to reduce transmission rates. Chapter 3 discusses mass hysteria and how people from all walks of life behaved in an irrational way throughout the pandemic and ended up supporting uncritically all pandemic policies. This is a key element in the discussion and it explains the fixed beliefs people had on the virus which paved the way for politicians to both make not well-grounded scientific claims and to implement the policies in question. Mass hysteria needs to be discussed after Chap. 2, for it is important to demonstrate first that the virus is much less deadly than it was officially claimed and then to describe the behavior of the population as irrational. It is moreover

1.1  Public Interest Explanations vs. Public Choice Explanations

5

important to place this discussion prior to Chap. 4, for the failure to consider the proper tradeoff is in part explained by the fact that the panicked people (either professors or grassroots individuals) ignored the costs of the lockdowns. Moreover, some assessments made by economists on the costs of lockdowns make sense mainly in the light of the prevailing mass hysteria. Chapter 4 puts into test the public interest claims in the domain of economics and of social sciences (summarized as prediction IV above). As noted, policy makers missed the fundamental tradeoff: the dilemma  was never lives against money lost, but lives that are threatened by the virus against lives that are threatened by the steep economic downturn. Estimating the proper tradeoff reveals that many years of life lost are about to occur as a result of the economic devastation inflicted by the lockdowns. The toll school closures took on children is also discussed as well as the effects of isolation in the population. I argue that to avoid that much damage, it would have been enough for policy makers to take into account widely disseminated and well-established knowledge from diverse fields of social sciences. Therefore, the main conclusion to be drawn from Chaps. 2, 3 and 4 is that the public interest claims of the pandemic policies are rejected both vis-à-vis medicine and with respect to social sciences and that not only they are incompatible with science but that they moreover caused serious harm. If predictions I–IV were accurate, the public choice theory could still find application, albeit the public choice theorist would have had a more difficult case to make than the one I am making in this book. For example, they would need to explore possible government malfeasance that protected one group more while exposing others to a death threat and to find evidence which could link such wrongdoing with pressure groups that directed optimal protection-policies in favor of some groups and at the expense of the lives of some others (the ones who were exposed to the virus). But if, by and large, millions of lives were saved through lockdowns and similar interventions, the public interest theory would have been largely vindicated and the public choice approach would have had limited applicability. However, by examining the available evidence and by rejecting the premises upon which the public interest narrative relied on, we have solid grounds to claim that the policies in place were not evidence-based and that the pandemic policy responses were not implemented to serve the public’s best interests. This in turn leaves much room for exploring a possible public choice mechanism that explains such decision-making. As public interest theorists of the pandemic policy responses should be considered first and foremost the politicians who implemented lockdowns, the rhetoric of whom was carefully designed through war-metaphors and sensationalization so that the necessity of these measures will appear as foolproof (Dada et  al., 2020). Secondarily, advocates of these measures that, apart from the general public, came moreover from Universities (Edmod et al., 2020; John Snow Memorandum, 2020), most media outlets in pro-lockdown countries as well as in no lockdown regions, and pundits who endorsed these policies and repeatedly doubled down on them. Since they all used pretty much the same arguments during the course of the pandemic, I use the umbrella term “public-interest theorists” to describe them, while when lockdowns are discussed in the text they are also described as “lockdowners” or “lockdown advocates.”

6

1 Introduction

Since the decision-making was not science-driven, we have to look for the motivations that led politicians to put in place such policies. And as Chap. 5 will demonstrate, public choice theory provides a fine-grained explanation of what has happened. The predictions of public choice theory in the framework of the Covid-19 pandemic can be summarized: ( A) The pandemic policy responses are unrelated to science. (B) The mitigation measures aim at benefiting those who implement them and not the majority of the population. (C) The majority of the population is overly fearful of the virus and supports the mitigation measures regardless of their costs and their effectiveness. (D) Politicians and bureaucrats are eventually better off, while huge segments of the population are worse off. The public was in a state of mass hysteria and was unable to rationally assess the risk of the virus and the dangers induced by the mitigation measures. This became a fixed belief in the society and it was the main element of peoples’ expected utilities. Thus, people demanded action from politicians, and in particular, they expected policies that would satisfy their well-established belief that this is an unprecedented health crisis and that all people face serious death threat. Politicians took the bull by the horns and implemented policies that could meet these demands while simultaneously benefiting greatly by this action by establishing their popularity and by gaining reelection. The side effects are horrendous, but at the time the interventions took place, very few raised reasonable concerns; the vast majority supported politicians and pressed for more and more of the same, not scientifically grounded, ineffective and harmful policies. To cut a long story short, almost everyone ended up worse off, except for politicians, who ended up better off.

References Arrow, K. (1963). Social choice and individual values. 2nd Edition. New York: Wiley. Buchanan, J. M., & Tullock, G. (1962). The calculus of consent: Logical foundations of constitutional democracy. University of Michigan Press. Dada, S., Ashworth, H.  C., Bewa, M.  J., & Dhat, R. (2020). Words matter: Political and gender analysis of speeches made by heads of government during the Covid-19 pandemic. BMJ Global Health. https://doi.org/10.1136/bmjgh-­2020-­003910 Downs, A. (1957). An economic theory of democracy. Harper & Row. Edmod, C., Hamilton, S., Holden, R., & Preston, B. (2020). An open letter by Australian economists on tradeoffs during the COVID-19 crisis. http://covid19openletter.net/. Accessed 19 Apr 2020. Hertog, d. J. (2012). Economic theories of regulation. In R.  J. Van den Bergh & A.  M. Pacces (Eds.), Regulation and economics. Edward Elgar Publishing. John Snow Memorandum. (2020). https://www.johnsnowmemo.com/ Jones, D. (1988). Regulatory concepts, propositions, and doctrines: Casualties and survivors. Journal of Economic Issues, 22(4), 108–190. Mitnick, B. M. (1980). The political economy of regulation. Columbia University Press.

References

7

Niskanen, W. A., Jr. (1971). Bureaucracy and representative government. Aldine-Atherton. Pigou, A. C. (1932). The economics of welfare. Macmillan and Co. Posner, R.  A. (1974). Theories of economic regulation. The Bell Journal of Economics and Management Science, 5(2), 335–358. Shughart, W., & Razzolini, L. (2001). The Elgar companion to public choice (2nd ed.). Edward Elgar. Stigler, G. (1971). The theory of economic regulation. Bell Journal of Economics and Management Science, 2(1), 3–21.

Chapter 2

Viral Mitigation: Weak Theoretical Underpinnings

Mitigation measures included primarily lockdowns and masks and, later in the pandemic, mass vaccination. All of them were supposed to eradicate the disease or at least “flatten the curve.” To stress the need for disease eradication and/or the need for reduced transmission rates, three postulates were put forward by the proponents of the pandemic policy responses. First, it was claimed that the virus poses a high death risk to all age-groups, and so we need policies that will be able to offer protection to all people. This is the first postulate, which I would like to call as the “equal vulnerability thesis.” Second, the claim that there is no pre-existing immunity and hence all people are equally susceptible to the virus, which is the “equal susceptibility thesis.” The third postulate is that the coronavirus can be transmitted not only by symptomatic but also by asymptomatic people. This is the “equal infectivity thesis.” These three premises were mistaken, and the pandemic policies, i.e., lockdowns, masks, and mass vaccination, failed to achieve their declared goals, i.e., they did not eradicate the disease and did not impact on transmission rates.

2.1 SARS-CoV-2’s IFR Speaks Against the Equal Vulnerability Thesis The case fatality rate (CFR) of a disease is the proportion of the number of people that have died among the number of people that have been identified as cases. Identified cases are the officially recorded ones. Back in March 2020, the World Health Organization (WHO) came up with a CFR estimate of 3.4% (WHO, 2020a) which is considered to be enormously high. But as it usually happens with respiratory diseases, like the novel coronavirus, the number of people that have in fact contracted the virus is much higher than the number of recorded cases. This implies that the CFR misses a huge portion of people that have contracted the virus, but they did not even realize it. In the case of SARS-CoV-2, it seems that more than one-third © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Karadimas, The Covid-19 Pandemic, Studies in Public Choice 42, https://doi.org/10.1007/978-3-031-24967-9_2

9

10

2  Viral Mitigation: Weak Theoretical Underpinnings

of infections are totally asymptomatic (Pratha et al., 2021). That is, they developed no symptoms or very mild ones and recovered speedily and well from it without asking for medical care. This is represented by the infection fatality rate (IFR) which is the proportion of deaths among all the people that have had the virus and have not been recorded officially as cases. So in order to assess how dangerous a disease is, one needs to know the IFR with the greatest possible precision. The best way to have an accurate estimate of the IFR is to measure the seroprevalence, i.e., the levels of antibodies in the society. The initial IFR estimate that became known to the public was by the WHO, and quite surprisingly, it was in the region of the CFR due to the assumption that there were very few, if any, asymptomatic carriers of the virus (WHO, 2020b). Mathematical modeling subsequently came into play by researchers at the Imperial College, and the IFR was reduced approximately to 1% and, in particular, to 0.9% (Ferguson et al., 2020). So the data that have so far been cited imply that roughly 1 in 30 people will die from Covid-19 according to the CFR and the IFR of the WHO and approximately 1 in 100 people given the estimate of 0.9% IFR of the Imperial College. However, as stated, the CFR is not to be regarded as a reliable measure of how dangerous a disease is, and hence, it should be ruled out. Thus, we turn to the IFR. Is the IFR of Sars-Cov-2 0.9% or is it 3.4%? These estimates were refuted shortly thereafter. Data from Iceland reduced significantly the IFR to 0.3% (Gudbjartsson et al., 2020), and several seroprevalence studies indicated that the IFR is much lower than the initial estimates suggested (Bendavid et  al., 2021b). Α revealing study by Ioannidis (2021) shows that the median IFR is 0.15%, i.e., approximately 1 out of 670 infected people die. More importantly, for people below 70 years old, the IFR is further reduced to 0.05%, i.e., 5 out of 10,000 infected people die. However, the likelihood of serious disease and death increases steeply among those above 70 years old, and it may go even up to 1.54% or even higher due to comorbidities (Ioannidis, 2020). It is possible that the current antibody studies underestimate the immune responses in the population and thus overestimate the IFR.  Serological studies do not account for the T-Cell responses that are either pre-existing or are elicited after mild or asymptomatic Covid-19, and they also are structured so that they detect IgB and IgM antibodies. They thus do not detect IgA antibodies that are also important in fighting pathogens and are also produced during infection. Those who tackle the disease through T-Cells solely or through IgA antibodies may not develop virus-specific IgG antibodies, and so the prevalence of the disease may be considered lower than it actually is. Moreover, even if IgG antibodies are secreted, they appear to decline rapidly, so late testing may miss some cases of these antibody responses too (Burgess et al., 2020). In spite of this downside, serological studies still give us a quite reliable estimate of how lethal a disease is, for even if they do not project with 100% precision the IFR by somehow underestimating the levels of immunity and slightly overestimating the IFR, they do give us a clear picture of who is at high risk and, speaking of Covid-19, of the astounding differences in mortality between the young and the

2.1  SARS-CoV-2’s IFR Speaks Against the Equal Vulnerability Thesis

11

old—which is what matters most when it comes to public policy. The approximate age-stratification of the IFR makes it clear.1 –– –– –– –– –– –– ––

For children aged 0–19 years old, it is 0.001%. For people aged 20–29 years old, it is 0.010%. For people aged 30–39 years old, it is 0.023%. For people aged 40–49 years old, it is 0.050%. For people aged 50–59 years old, it is 0.15%. For people aged 60–69 years old, it is 0.49%. For people over 70, the range is well above 1% and may go even higher than 4.9%.

Therefore, if 100,000 children are infected, it is estimated that approximately 1 will die, and if the same number of infections occurs in people over 70, then we could expect even 4000 of them to die. If the same number of infections occurs primarily to people over 80, the IFR will further increase. So, if the elderly are exposed to the virus, then the death rates are likely to be very high, but if the youngsters are exposed and the elderly are exposed less, then the death rates could end up being very low. If several age-groups are exposed randomly, then the data suggest that the median IFR would be between 0.1% and 0.4% (closer to 0.1% if the youngsters are more exposed and the elderly are protected and closer to 0.4% if the converse happens). So, the appropriate answer to the above question is that the IFR of Sars-Cov-2 is not a constant and is related to the age-group of people that are getting infected. The median age of people who died of Covid-19 in several countries across the globe vindicates one more time that Covid-19 poses high risk of death to septuagenarians and octogenarians and not to everybody. In Australia, the median age of Covid-19 deaths is 82 (Australian Government Department of Health, 2021), while the life expectancy is 83.64 (Macrotrends, 2021). In Belgium, the median age of virus-induced deaths is 86 (Sierra et al., 2020) and the lifespan is 80.8 (For a Healthy Belgium, 2021). In South Africa, these figures are 67 and 64.38, respectively (SAcoronavirus, 2021); in the UK, 83 and 81 (ONS, 2021a), and in the USA, 78 (CDC, 2022b) and 77 (CDC, 2022c). As one can see, the median age of Covid-19 deaths is in the same range as the average lifespan, if it does not surpass it on some occasions. Lockdowners could object that I wrongly use the median term and that it could be more appropriate to use the mean estimate which calculates the average age of deaths involving or due to Covid-19 and not the middle number as the median term does. First and foremost, even if the above numbers involved the mean instead of the

 These estimates are drawn from an update by Axfors and Ioannidis (2022). Those interested in the data should check the “Appendix Text 1” of this paper because the main body of the paper focuses on IFRs for the elderly (either community or non-community dwelling). The Centers for Disease Prevention and Control has given IFRs (CDC, 2020c) which verify that there is apparent agreement in the literature that, even if the median IFR slightly differs from country to country, it was very early on known that there is a huge difference in mortality rates between those over 65–70 and those below this age. 1

12

2  Viral Mitigation: Weak Theoretical Underpinnings

median estimate, the central conclusion is the same, for the differences in numbers would not be so great as to challenge the fact that Covid-19 is a death threat for people over 70 and not for everyone. To back this up, it is, I think, enough to mention that the mean age of deaths involving or due to Covid-19 in the UK is 80.4 (ONS, 2021a), which is in the same region with the average lifespan (81) and similar is the case with other countries. Most importantly, the median is more accurate when it comes to estimate age-related death rates from a particular disease because it is robust against outliers, whereas the mean term is not. Outlier parameters can skew our estimates. For example, a person aged 105 years old may have died of Covid-19, and if we measure the mean age and take this estimate into account, then the mean age of deaths may go upwards and conversely it can go downwards if, say, a baby dies of Covid-19 (which is extremely unlikely but one cannot exclude it formally). On the contrary, by measuring the median age of deaths, we arrange the values from the lowest to the highest and take the one that is in the middle (if the overall number of values is even, then the median is the mean value of the two values that are in the middle). So, large fluctuations are avoided. While age is the main criterion to determine who is at risk, some serious co-­ morbidities also need to be taken into account to have a complete picture of the segment of the population that is at risk. A meta-analysis examined 76 studies across 14 countries to determine the groups that could be more threatened by the virus and found that people over 75 years old, individuals with severe obesity, and individuals with active cancer are the three groups that could suffer severe clinical outcomes if they contract the virus (Booth et al., 2021). Age (primarily) and serious underlying health conditions (secondarily) like those mentioned are the drivers of hospitalization and death irrespective of the appearance of Sars-Cov-2, and there is evidence indicating that the relative all-cause mortality risks of the vulnerable in comparison to the general population were not dramatically altered upon the arrival of the novel coronavirus. On the contrary, they remained stable during 2020 compared to 2015–2019. A big Swedish study demonstrated exactly this. Nationwide all-cause mortality between March and September 2015 measured against the first reported Covid-19 wave in March through September 2020 shows that the death rates of people with rheumatoid arthritis were proportional to the general population (relative risks around 2 and 1.5, respectively, that were not higher during 2020 than during 2015–2019), and the hospitalization rate was fairly low too (0.5%) (Bower et al., 2021). Thus Covid-19 did not increase the absolute risks for the majority of the society, including people with underlying conditions such as rheumatoid arthritis, and that it only threatened roughly those over 70, the severely obese and those with active cancer. It speaks volumes that the credibility of the equal vulnerability thesis vanishes. Or, at least, this ought to have been the case in March to April 2020 when the first data from serological studies started pouring in. However, policy-makers totally ignored these estimates and kept repeating that all age-groups face pretty much the same risk of dying, and thus a lockdown is needed in order to protect all of them. An alternative strategy, which was in accordance with the available evidence, was the so-called “Focused Protection.” Renowned epidemiologists and public health

2.2  There Is More to Immunity Than Antibodies: The Case of Pre-existing Immunity

13

scientists, such as Jay Bhattacharya, Sunetra Gupta, and Martin Kulldorf, offered an alternative to lockdowns by claiming that since there is more than a thousand-fold difference in mortality between the youngsters and the elderly, the mitigation measures should also be age-focused namely, we should let the former go about their lives and take targeted measures to protect those at really high risk (Great Barrington Declaration, 2020). Since the median IFR for those under 70 is in the region of 0.05% and for those over 70 well above 1%, it is clear that to claim that all people face the same risk of dying is not an evidence-based assessment and to implement lockdowns based on this same claim is not an evidence-based strategy. Thus, the equal vulnerability thesis is not backed up by science, and lockdowns were not the result of an evidence-based attempt to protect all people.

2.2 There Is More to Immunity Than Antibodies: The Case of Pre-existing Immunity An important scientific finding that policy-makers totally missed was the fact of pre-existing immunity in the population which suggests that many a people were not even susceptible to Sars-Cov-2. On the contrary, in the early days of the pandemic, it was repeatedly emphasized that since no one had contracted the virus, no one had antibodies, and thus there was no immunity at all. After a few months, when a chunk of the population was infected, it was said that only those who had detectable antibodies had immunity. Once the antibodies wane, there is no immunity and we are back to square one. We are thus on the verge of an endless pandemic. This argument was based on a gross misunderstanding regarding the very notion of immunity. To appreciate why pre-existing immunity is important, we should first present a few things about how our immune systems work. This will make clear that when it comes to viral infections, antibodies are helpful but are not the game-­ changer many may think they are and that our immune systems have cellular immune mechanisms that are at work which can protect us from numerous pathogens regardless of whether antibodies against a particular pathogen circulate in the blood stream. Our immune system is structured as a two-tiered defense mechanism. The one level involves innate immunity and the other adaptive immunity. Innate immunity offers protection against any pathogen that may enter the body. From germs that can cause respiratory diseases to bacteria entering through a wound in the skin, innate immune responses try to detect and destroy or kill the foreign substance on the spot. It is often called “non-specific” immunity since it does not confer memory immunity to the host (Mak & Saunders, 2006). When the innate immune system does not manage to stop the intruder, then it is adaptive immunity that takes over. Adaptive immunity offers specific immune response to certain pathogens, and it is composed of B-Cells and T-Cells. When a novel, i.e., not previously encountered, germ enters and innate immunity does not kill it, adaptive immunity identifies it and then kills it.

14

2  Viral Mitigation: Weak Theoretical Underpinnings

It takes some time to identify the novel pathogen in the first place, so adaptive immunity is slower than the innate response at this stage, but it is more accurate than that, and upon re-exposure, it acts instantly. The second infection from the same pathogen therefore is either entirely without symptoms or with only mild symptoms. B-Cells and T-Cells both contribute to clear the disease. B-Cells are responsible for the humoral adaptive immunity, while T-Cells set in motion cellular adaptive immunity. They recognize molecular components of the germ, known as antigens. The recognition of these antigens is done by specific receptors present in the cell surface of B-Cells and T-Cells. The way B-Cells and T-Cells work to recognize the antigens differs greatly. B-Cells produce huge Y-shaped proteins, the well-­ known antibodies, which are specific to the pathogen and have the ability to bind on the antigens and neutralize them or label them for destruction. The antigen portion binding to the antibody is called “B-Cell epitope.” T-cells, on the contrary, present on their surface a specific receptor known as the “T-cell receptor” that enables the recognition of antigens when they are displayed on the surface of antigen-­presenting cells (APCs) bound to molecules known as the “major histocompatibility complex” (MHC). MHC molecules are recognized by two distinct subsets of T-cells, CD8 and CD4 T-cells. There are therefore CD8 and CD4 T-Cell epitopes. T-Cell epitopes are peptides derived from antigens2 and recognized by the T-cell receptor (TCR) when bound to MHC molecules displayed on the cell surface of APCs. CD4 and CD8 T-cells recognize different MHC-peptides and start work against the pathogen in various ways. The important thing is that T-Cell activation not only helps eliminating the infection, but it also paves the way for doing that more quickly when the immune system is re-exposed to a specific invading antigen (Tricado et al., 2017). This discussion makes clear that while antibodies secreted by the B-Cells are important in fighting the infection by neutralizing the pathogen, they are far from being the only line of defense we have. In fact, the more robust immune response is the one induced by the T-Cells, which target a much greater number of epitopes of the pathogen than antibodies. So when a pathogen, a virus in the case of Sars-Cov-2, mutates to evade antibodies, it is unlikely to mutate to evade T-Cells. Virus-specific T-Cells persist for years. A notable result of this sort emerged from Sars-Cov-1 that circulated back in 2003. Researchers found specific T-Cell immunity even 17 years after the infection (Le Bert et al., 2021). So even if the population was totally immunologically naïve against Sars-Cov-2, i.e., there was zero memory against it, serious cases of reinfections, let alone endless repeated reinfections, were an implausible scenario for the simple reason that it was known that T-Cells would not allow this to happen. But some important results came to the fore early on in the pandemic which should have made us even more optimistic about the pandemic trajectory. There was robust T-Cell immunity in the population even prior to the advent of the virus which suggests that a chunk of the population was already immune to Sars-Cov-2. In particular, a range of pre-existing memory CD4 T-Cells that are cross-reactive with

 Thus, peptides are proteins that are broken up into amino acids.

2

2.2  There Is More to Immunity Than Antibodies: The Case of Pre-existing Immunity

15

comparable affinity to SARS-CoV-2 and the common cold human coronaviruses (Hcovs), HCoV-OC43, HCoV-229E, HCoV-NL63, and HCoV-HKU1 were present (Grifoni et al., 2020; Mateus et al., 2020). Scientists call them pre-existing because they are detected in individuals with no known exposure to Sars-Cov-2 and especially in seronegative ones, i.e., people with no detectable antibodies in their blood (Sekine et al., 2020).3 Even though policy-makers and lockdown advocates took this to be a trivial finding, it was not. It debunks one of the main pro-lockdown claims, according to which, since this is a new virus, there is zero immunity against it and so everyone is equally susceptible. Sars-Cov-2 may have been a new virus, but it does not follow that there is no immunity against it. This was known in theory since innate immunity guarantees that there will be some reaction against any possible pathogen, but it was suggested even more emphatically, by the way T-Cells work. A recent example, prior to the Covid-19 pandemic, was pre-existing T-Cell immunity against H1N1 (Sridhar et al., 2013)4 which appeared in 2009 when the WHO had also declared a pandemic (WHO, 2009). The H1N1 influenza strain was similar and comparable to influenza strains that circulated 50–60 years prior to 2009, and this led to increased levels of pre-existing immunity against H1N1, especially among the elderly (Skountzou et al., 2010). This precedent should have made lockdowners more reluctant to claim that equal susceptibility holds when a novel pathogen appears. T-Cells’ ability to confer long-lasting memory immunity against a great number of epitopes implies that for a pathogen to evade T-Cells, it should have to present a completely new set of proteins, and, speaking of coronaviruses, the family of viruses to which Sars-Cov-2 belongs, that is not very likely since they mutate rather slowly and thus there is a great deal of similarities between the different strains. So, since people co-exist with coronaviruses for centuries, it was not an implausibility that some cross-reactive immunity would appear. The robust pre-­ existing T-Cell immunity mechanisms vindicated that.5 The fact that the majority of cases had no symptoms or only mild ones means that the immune system cleared the infection rather quickly due to instant response. My takeaway from this discussion is that if no memory against Sars-Cov-2 existed, the asymptomatic and mild cases would have been fewer for it would take more time for adaptive  Later in the pandemic, in December 2020, it was shown that apart from T-Cell pre-existing immunity, there is also antibody cross reactive protection. That is, pre-existing antibodies specific to Sars-Cov-2 were identified (Ng et al., 2020), which seems to totally demolish the equal susceptibility thesis, for it suggests that even if our immune systems were based solely on antibodies, again not all people would be susceptible to Sars-Cov-2. 4  The CDC has also offered an analysis on the issue and also documented that among people over 60 years old, pre-existing immunity to H1N1 was in the region of 33% (CDC, 2009). 5  It becomes therefore clear that there was literally zero science behind the panic over variants and the mitigation measures that followed, for the simple reason that, in case of re-exposure, T-Cell Immunity is able to block Sars-Cov-2 variants with remarkable ease, and thus those who claimed that variants such as Delta or Omicron would transmit as if they were new viruses were simply mistaken. Since, unsurprisingly, there are studies that have proved empirically that variants do not overcome naturally acquired immunity (Tarke et al., 2021; Chemaitelly et al., 2021; Abu-Raddad et al., 2021; Altarawneh et al., 2022), and I am not going to discuss the issue further. 3

16

2  Viral Mitigation: Weak Theoretical Underpinnings

immunity to activate B-Cell and T-Cell responses, and in the meantime, the agent would have suffered more symptoms, though without necessarily making the virus more deadly. In other words, for those with memory CD4 T-Cells, the initial exposure to Sars-­Cov-­2 appears to have been equal, in terms of severity, to a reinfection for those who had no CD4 T-Cells when exposed in the first place to the virus. Policy-makers behaved as if no pre-existing immunity exists and shaped policy-­ making based on this mistake. Prior to declaring zero immunity and equal susceptibility against Sars-Cov-2, the pundits and the governments should have taken some basic facts about our immune responses into account. By insisting on the equal-­ susceptibility thesis even after pre-existing immunity became a well-documented fact, their claim becomes totally erroneous. Of course, this was not the last mistaken premise used to justify mitigation measures.

2.3 Is There Such a Thing as Asymptomatic Transmission? During the Covid-19 crisis, an asymptomatic carrier of Sars-Cov-2 was considered to be a person tested positive in a PCR test. But the PCR test does not distinguish between infectious virus and non-infectious nucleic acid, and for many viral diseases (Human Coronaviruses, Sars-Cov, Middle East Respiratory Syndrom, Influenza Virus, Ebola virus, and Zika virus), it is known that viral load can be detected long after the disappearance of the infective viral load (Atkinson & Pedersen, 2020). The PCR test identifies fragments of the virus, which are many times dead, while for transmission to occur completely, live viruses are required. So a proper interpretation of such an RT-PCR test requires consideration of patient characteristics such as symptoms and their severity, contacts history, presence of pre-existing morbidities and drug history, the threshold value, the number of days from symptom onset to test, and the specimen donor’s age. Overall, those with high value threshold are unlikely to have infectious potential (Jefferson et al., 2020b). It is therefore far from clear that a positive PCR test means infection from Sars-Cov-2, let alone viral transmissibility (Jefferson et  al., 2020a; Mina et  al., 2021; Stang et al., 2021). Untrammeled by this glitch, lockdown advocates claimed emphatically that asymptomatic transmission is ubiquitous and thus all people should be considered as infected persons that can transmit the virus. So we ended up to implementing a strategy aberration: instead of suggesting isolation to the symptomatic, the healthy people were quarantined on the assumption that they may carry a deadly pathogen which they could transmit it without realizing it. If we do not quarantine everyone, the argument goes, cases will spiral out of control, and this, in combination with the inflated IFRs and the equal susceptibility thesis, will result to an unconscionably high death rate. This line of reasoning was never in the past adopted by scientists and/or health authorities, and it would have been absurd even if the PCR test was totally reliable. The standard strategy was that no matter how deadly a pathogen has been, quarantines are to be suggested (or enforced) to those who developed

2.3  Is There Such a Thing as Asymptomatic Transmission?

17

symptoms of the disease. The WHO, for example, suggested, at least until 2019, that in order to best address an influenza pandemic, only the symptomatic people should be isolated until they recover and not even those that came in close contact with them, let alone those that were not even exposed to an infected person (WHO, 2019). There is a possibility that asymptomatic people may shed influenza virus, but the viral load is so small that is unable to cause infection to others. No evidence was ever available to show that asymptomatic or pre-symptomatic transmission of influenza viruses takes place. Thus, asymptomatic transmission of respiratory illnesses was considered a minor and limited exception—if we accept that it ever happens at all—to the general rule according to which transmission occurs through virus-laden respiratory droplets and small particle aerosols emitted by symptomatic individuals (Patrozou & Mermel, 2009). Indeed, when Sars-Cov-2 started circulating, no reliable data suggested that asymptomatic transmission was the rule. The best lockdown supporters were able to show a flawed publication in the New England Journal of Medicine (NEJM) in January 30, 2020. Authors of this piece reported that a businesswoman from Shanghai who had no symptoms met with several people in Germany and that this led to a cluster of infections. The authors wrote that the woman developed symptoms on her flight back to China. Based on this story, the authors confidently concluded that Sars-Cov-2 can be transmitted asymptomatically and called for a possible reassessment of transmission dynamics (Rothe et al., 2020). Even if that were the case, it is dubious how we can throw out of the window all our knowledge on the transmission of respiratory diseases, take a report based on a single case at face value, and conclude that asymptomatic transmission is frequently occurring. The case for asymptomatic transmission could have been reconsidered if a robust randomized control trial (RCT) was conducted or a well-worked presentation of data from several settings in which a coronavirus outbreak occurred and which would show that asymptomatic transmission of Sars-Cov-2 is significantly more frequent than we already knew it has been for other coronaviruses or flu-viruses. Taking a narrative like the one published by the NEJM as gospel was methodologically problematic even to one with no formal training in epidemiology, so one is right to wonder how on earth, based solely on this note, asymptomatic transmission almost became the new scientific consensus on transmission and faced little to no criticism. But things are even worse than that. The story was mistakenly reported because the authors—as they admitted—did not talk to the woman prior to sending the letter. Contrary to what the authors contended in the paper, the woman was already sick when she was in Germany and met with her colleagues (Kupferschmidt, 2020). Thus, she infected them because they met while she had developed symptoms of the disease. The kind of transmission therefore that occurred was the standard symptomatic transmission that is possible to happen when one meets with a sick person. So, as if it were not on its own highly questionable to claim that asymptomatic transmission is common based on a case reported in a narrative form, this inference was not even based on a true story! Science on coronaviruses has not changed on this issue, and no sound findings indicating that the novel coronavirus is transmitted asymptomatically have been

18

2  Viral Mitigation: Weak Theoretical Underpinnings

brought to the fore so far. On the contrary, there is evidence which show that asymptomatic Sars-Cov-2 transmission hardly ever happens, a result which seems to verify the studies on influenza transmission which, as we saw, also show that no clear-cut evidence of asymptomatic transmission is on offer. Researchers examined 455 individuals who came into close contact with an asymptomatic virus carrier, and none of the 455 agents was considered to have been infected from the asymptomatic individual (Gao et al., 2020). A large study from Wuhan in China reported that amongst 1174 close contacts of 300 asymptomatic individuals, no infection occurred. All asymptomatic positive cases, re-positive cases, and their close contacts were isolated for at least two weeks until the results of nucleic acid testing were negative. None of the detected positive cases or their close contacts became symptomatic or newly confirmed with Covid-19 during the isolation period (Cao et al., 2020). In view of the preceding, it should be surprising, to say the least, that so many people endorsed the thesis that asymptomatic transmission is the driver of the pandemic, and based on this, they argued in favor of lockdowns. Since it is not backed up by science, the claim that asymptomatic or pre-symptomatic people may, unbeknownst to them, transmit the virus, it appears that it is not an overstatement to claim that asymptomatic persons are in fact healthy persons, and thus, decision-­ making that quarantines the entire population, including healthy individuals, makes little sense from a scientific point of view. The reasons that asymptomatic transmission was taken to heart will become clearer in the following chapters when mass hysteria is discussed. My conviction is that mass hysteria was at its heyday when the wrong study published by the NEJM was publicly discussed and media coverage on this topic convinced almost everybody that they may get infected by simply talking to each other or by shaking hands with each other. Then lockdowns ensued and this claim was consolidated for it is as if politicians implied that if no lockdowns were enforced, asymptomatic transmission would thrive. It was thus not an easy task to present the evidence regarding the routes of transmission and to convince the majority of people that asymptomatic transmission is, if not a fallacy, then an extremely rare state of affairs.

2.4 Wishful Thinking Part 1: Zero-Covid Through Lockdowns The rationale behind the lockdowns and other mitigation measures, such as masks, was that they can lead us either to a “zero-covid” world or, at least, they would “flatten the curve” so that cases will not spike exponentially and optimal treatment for most people would be available. This is possible, it was claimed, if people change their behavior and interact less with each other. Thus, all people should stay at home, work from home, and study from home. One possible source of inspiration for this assertion has perhaps been the fact that since other type of diseases, such as

2.4  Wishful Thinking Part 1: Zero-Covid Through Lockdowns

19

Aids, can be mitigated through behavioral changes (d’Onofrio & Manfredi, 2009), the same should be the case for a respiratory disease. We have seen in the previous sections that the postulates that gave a scientific veneer to lockdowns were mistaken, and we will now go on to see how lockdowns fell short of attaining their declared goals. One of the targets of lockdown was to eradicate the disease (Varghese, 2020; UK Parliament and Government Petition, 2021; Ndam, 2020). According to this view, if the lockdown is strict enough and people do not fail to adhere to the rules, then a particular country will eliminate the disease. If all the countries manage to do the same, then the whole world will be covid-free. However, with the exception of smallpox, no other infectious disease has ever been eradicated (WHO Smallpox) for humans (rinderpest has been eradicated from ruminants), despite the fact that there are several diseases that have been targeted for eradication, such as tuberculosis and polio. These eradication projects often combine pharmaceutical interventions, such as mass vaccination programs, and non-pharmaceutical interventions, such as protocols to be followed, but they have not managed to eradicate any other disease. Since smallpox eradication is the exception and not the rule, the claim that we will shut down the society in order to eradicate Sars-Cov-2 was never well grounded. Neither was the idea that this can be done by vaccinating the entire population. To discuss in detail the issue of disease eradication, in general, would probably require another monograph on its own right, but we can point out the main reasons that make the eradication of Covid-19  in particular not an achievable goal and that this disease is not to be compared with smallpox, which in turn shows that the feigned goal of eradication bears little to no scientific grounding and should have never been used as an argument to justify lockdowns. There are several problems besetting the “zero-covid” approach. The case for zero-covid collapses primarily upon the criteria the Task Force on Disease Eradication has set for a disease to be considered eradicable. There are three biological and four social criteria, and failure to meet some of them makes a disease to be considered ineradicable. As for the biological criteria, first is the need for effective and practical interventions that are likely to achieve eradication; second is the demonstrated feasibility of eradication; third is the “epidemiological vulnerability.” There are four other important social criteria: First, a broad social perception of the importance of the disease, second, a reasonable projected cost, third, synergy with other health system activities, and fourth, necessity for eradication rather than control (Hinman & Hopkins, 1998). Coronaviruses spread worldwide, become endemic, and are controlled by herd immunity. Below, we will elaborate a bit on the difference between endemicity and eradication, but for now, it suffices to stress that the eradication of a disease that has spread worldwide and has become endemic is not a disease with demonstrated feasibility of eradication since there are numerous infective agents (Barrett, 2004; Hall, 2011). The idea that a lockdown in particular will eradicate the disease collapses further upon the first biological criterion, since lockdowns were never used in order to control a disease, let alone to eradicate it, and so the assumption that lockdowns will be effective in eradicating Sars-Cov-2 was extremely dubious from the outset, and furthermore, it was known even prior to

20

2  Viral Mitigation: Weak Theoretical Underpinnings

their implementation that lockdowns were not a practical solution. When the second social criterion is considered, it becomes even clearer that the projected cost is unconscionable since lockdowns are an extremely costly strategy. Finally, the discussion on the age-stratification of the Covid-19 IFRs casts further doubt on the view that eradication instead of control of the disease is required. Although the failure to meet these criteria could have been enough to render the disease ineradicable, there are more to be said to reject the claim that zero-covid is possible by any means, let alone through lockdowns. Smallpox eradication, for example, was not the result of a lockdown, and lockdown supporters should have been aware of this before claiming that by shutting everything down, we will get rid of Sars-Cov-2. What happened in Peru provides a clear rebuttal to this pro-lockdown claim. The first reported case in Peru was on the 6th of March 2020. The government imposed a nationwide lockdown on the 15th of March 2020, just nine days after the first recorded case. The lockdown in Peru was perhaps the strictest in the world. The military was patrolling the streets; an 8 pm curfew was imposed, and only men were allowed to leave home some days a week and only women some others (Wikipedia, 2022). If strict lockdowns were supposed to eradicate the disease, then Peru should have been a covid-free champion. However, cases in Peru exploded after the lockdown was implemented, and Peru faced one of the highest death rates in the world. Indeed, if Peru’s case shows something, it is that even a very strict lockdown that is enforced very early on is unable not only to eradicate the coronavirus but also to reduce transmission rates. A closer look at the data demonstrates exactly this. On the 15th of March 2020, when the lockdown was enforced, Peru had only 0.16 cases per million population (pmp), and after the enforcement of the lockdown, cases increased rapidly and they reached a peak in 2nd of June 2020 (207.33 pmp) and peaked once more on August 19, 2020 (255.40). During the second covid-wave in Peru, cases peaked twice; on March 29, 2021 (286.04) and on April 13, 2021 (297.62) (OWD, 2021c). Thus, the lockdown in Peru was not only unable to eradicate the disease, but Peru also faced a repeated spike in cases while the strictest possible lockdown was in place. I will offer an explanation of the latter below in this chapter, but for the time being, it is important to stick to the inability of lockdowns to eradicate the disease no matter how strictly they are enforced and how long they are in place. New Zealand is a country that is often touted as a paragon of a zero-covid state. However, the eradication-through-lockdown project failed there too. The failure there is even more blatant because New Zealand is an island, which suggests that border closures are more likely to provide some effect at least for some time. New Zealand introduced a strict lockdown on March 25, 2020, and lifted it on June 8, 2020 (Unite Against Covid-19, 2020), albeit by retaining strict border closures in place. When the first lockdown was lifted, there were zero recorded cases in New Zealand. In spite of this, and despite the fact that tight border controls were still enforced, cases appeared once more on 8th of August 2020 (OWD, 2021d). In a similar vein to New Zealand, Australia enforced repeated strict lockdowns and also appeared as a candidate for the zero-covid grant, albeit unsuccessfully. Notwithstanding the very strict measures Australian government enforced, which

2.5  Wishful Thinking Part 2: Zero-Covid Through Mass Vaccination

21

were in place for many months, the virus was not eliminated from the country, and a spike in cases and deaths was reported in August 2021 (OWD, 2021e). The cases of Peru, New Zealand, and Australia make it clear that lockdowns are not a mechanism for disease eradication, and even if they are enforced early on, stay in place for many months, are very strict, and are re-imposed again and again with the same high rates of stringency, they will fail to eradicate the disease. Never in the history of mankind have worldwide lockdowns been implemented, and never in the past has a disease been eradicated through lockdowns. Therefore, one would go as far as to say that those who argued that eradication through lockdowns is possible either gravely misunderstood the issue or pursued personal goals which were unrelated to the issue of actually addressing the pandemic.

2.5 Wishful Thinking Part 2: Zero-Covid Through Mass Vaccination The case for zero-covid appeared anew upon vaccine arrival (Wilson et al., 2021). Advocates of the idea suggested that we should vaccinate the entire planet so that we will eliminate the disease. There is evidence suggesting that vaccines provide some short term (4–6 months approximately) protection against hospitalization for some age-groups,6 and this rapidly waning immunity should have been enough to indicate that covid eradication is not achievable through vaccination. Moreover, we should be heedful of the inferences regarding vaccine effectiveness in the first place because the claim for zero-covid-through-mass-vaccination was assumed to be initially based on the data released prior to the commencement of mass vaccination. However, to the best of my knowledge, no such data were ever available. Moderna declared 100% (!) effectiveness against severe disease (Moderna, 2020), and as for Pfizer’s vaccine (BNT162b2), which claims 95% effectiveness, the available sources for its success when the vaccine was first rolled out were its own report (Pfizer, 2020), the FDA’s report on it (FDA Briefing Report, 2020), and a NEJM publication by Polack et al. (2020). Speaking of Moderna’s claim, this is not the place to evaluate the assumed 100% effectiveness against severe disease, since protection against severe outcomes is not the same as protection against infection, and when it comes to eradication, it is protection against infection that matters. Hence, Moderna never claimed that its vaccine is able to protect against infection. As for Pfizer’s vaccine, the early reports just mentioned seemed to indicate that the vaccine reduces the disease incidence, for they suggest that fewer people in the vaccinated group developed Covid-19 in contradistinction to the control group. However, if the plan is to eradicate a disease, we need no one to develop symptoms related to this disease, and

 The CDC suggested four months protection of mRNA and viral vector vaccines (CDC, 2021b), while another study (Tartof et al., 2021) suggested six month protection after mRNA vaccination. 6

22

2  Viral Mitigation: Weak Theoretical Underpinnings

the data from trials did not offer such information.7 Thus, the claim that it is possible to vaccinate all individuals in all the countries in the world so that the novel coronavirus will be eradicated was not elicited from these reports. It appears therefore that it was clear from the outset of mass vaccination program that the most important issue with covid vaccines is that vaccinated people can still catch the virus. As it quickly became known, the vaccines were not designed to test whether transmission rates are reduced (Doshi, 2020b), because as manufacturers themselves claimed, to conduct such an experiment would require swabbing people every two weeks to see whether transmission is blocked, which is, understandably, not an easily attainable project. Unsurprisingly therefore, the CDC released data in July 2021 showing that vaccinated and unvaccinated people carry the same viral load (CDC, 2021a), and thus, accordingly, they are equally likely to spread the disease. Further studies verified that the spread of Covid-19 is unrelated to the levels of vaccination in each region (Subramanian & Kumar, 2021), and it became increasingly clear that covid-vaccines do not stop transmission.8 The inability of the vaccines to stop the spread would have been enough to lead us to the conclusion that zero-covid through vaccination is not to be considered as a plausible scenario, and by furthermore taking into account the rapidly waning immune protection as well, the zero-covid thesis ought to have been no more in the limelight. However, one of the central claims was that we should take advantage of the vaccines, vaccinate everyone, and achieve eradication. The available data show clearly that there is no sound scientific basis to that claim, as there was no sound scientific basis to the claim that lockdowns will eradicate the disease. Both claims hinge on the same mistaken assumption, i.e., that we cannot live with a virus that poses a threat to  But that is not the only drawback. As the FDA’s report writes, there were 3410 total cases of suspected, but unconfirmed Covid-19 in the overall study population among which 1594 occurred in the vaccine group and 1816 in the placebo group. The phrase “suspected Covid-19” is definitely a source of confusion since it refers to symptoms akin to the ones caused by Sars-Cov-2, but which are not PCR-confirmed. Peter Doshi, editor of the prestigious The British Medical Journal (BMJ), notes that “suspected Covid-19” appears to be a category of disease that cannot be ignored, simply because it was not PCR-confirmed. He notes that if these “suspected Covid-19 cases” were the result of a false negative PCR test, this would further decrease vaccine efficacy. But mild flu-like symptoms, such as “suspected Covid-19” ones, can be caused by numerous pathogens: influenza strains, coronaviruses, rhinoviruses, adenoviruses, etc., and so these “suspected Covid-19” cases may be caused by other than Sars-Cov-2 pathogens. If suspected and confirmed Covid-19 cases had the same clinical outcome, then the analysis of the clinical course irrespective of etiology is required to assess whether vaccines will help reduce morbidity and the data released at the outset did not provide such information (Doshi, 2021). In view of this confusion, the claim that zerocovid is possible via vaccination becomes even weaker. 8  All the vaccines were based on the same mechanism. Genetic information is injected and the purpose is to enable the cells of the body to create the virus’s spike protein. The difference is that the mRNA vaccines rely on tiny lipids to get the genetic information into the cells while the viral vector ones have the genetic information carried into a dead virus (usually adenovirus) which transmits the information into the cell so that the latter would produce Sars-Cov-2’s spike protein. It was assumed that the immune system will react against this spike protein, and it will be “prepared” to fight the wild virus. All are prone to pretty much the same problems, i.e., they do not stop transmission, and thus this criticism applies to all of them at once. 7

2.5  Wishful Thinking Part 2: Zero-Covid Through Mass Vaccination

23

everyone and that we should take action to protect ourselves by either quarantining everyone and/or vaccinating everyone. A zero-covid-through-vaccination strategy would have been an onerous project and also an unlikely outcome even if the vaccines did stop transmission, but now it appears that it was nothing but a pious hope. In my view, the case for zero-covid, either through lockdowns or through mass vaccination, is a corollary of mass hysteria, and it has been exploited by politicians to pursue their own interests. Politicians misguided the public both on the issue of transmission and on the issue of vaccine-effectiveness. The public choice perspective on vaccines will be discussed in Chap. 5 and more will be said on how effective vaccines are against severe disease and what conclusions could be drawn when the immunity they confer is compared to naturally acquired immunity; since here we deal with the supposed mitigation of the disease, I will analyze a bit more the reasons that in my view make it clear that mass covid-vaccination was never possible to eradicate Covid-19. First, it is, I think, important not to confuse the notions of endemicity, which is the result of achieving herd immunity threshold,9 and that of disease-eradication. The former is definitely a plausible scenario in the case of Sars-Cov-2, and it appears to be already happening in some places, while the latter is not. Speaking of Sars-­ Cov-­2, it is commonly stated that the virus will eventually become endemic, and like all other Hocvs, which, as already cited, include OC43, 229E, NL63, and HKU1, it will be controlled by herd immunity. Once herd immunity threshold is achieved, people will contract Sars-Cov-2, but since most of them have immunity, they will not even realize it and only a tiny fraction of the population will become seriously ill—again, as it happens with other coronaviruses (Baillie et al., 2021). It should be noted here that the fact that human coronaviruses are considered to be mostly harmless, this is due to the naturally acquired immunity which persists in the population and which makes it difficult for these viruses to infect high-risk groups and is not the result of some inherent difference between human coronaviruses and Sars-Cov-2. In other words, all four Hcovs can also cause severe pneumonia to some old and immune-compromised people in the same way Sars-Cov-2 did. For example, an OC43 outbreak in care facilities in Canada in 2003 resulted in having 95 out of 143 residents and 53 out of 160 members of the staff catching the bug. Among the 95 residents, 8 died, 6 of whom after developing pneumonia. None of the staff members died nor suffered pneumonia (Patrick et al., 2006).10 Eight deaths out of 95 cases suggest an IFR and a CFR in the region of 8.4% which is similar, or perhaps could be even higher than the CFR and the IFR of Sars-Cov-2  in these

 It is not clear which is the herd immunity threshold for Sars-Cov-2. An interesting study (Aguas et  al., 2020), albeit, strangely, having not been peer reviewed, suggests that the herd immunity threshold may be in the region of 30% or even lower. Whatever the case, I will not have an eye for detail on that here, since it does not matter that much for our discussion. Either it is below 30%, above that, or exactly 30%, the outcome will be the same: Sars-Cov-2 endemicity. 10  An important finding of this study is that cross-reactive immunity between OCD43 and SarsCov-1 was identified, which suggests one more time that the equal susceptibility thesis was never scientifically robust, or at least not as robust as the politicians and the media told us it was. 9

24

2  Viral Mitigation: Weak Theoretical Underpinnings

age-­groups.11 The reason Sars-Cov-2 caused remarkable morbidity to the elderly in some places during 2020 was due to the fact that it was not endemic, combined with the lack of high levels of pre-existing immunity to the general population in these places, accompanied by a public policy failure to protect the elderly. Once the virus becomes endemic, it will become equally difficult for Sars-Cov-2 to infect the elderly as it now is for Hcovs to do so. The bottom line is that naturally acquired immunity and/or vaccine-induced immunity12 leads us from a state of worldwide viral spread to a state of endemicity where the virus still circulates but does not cause much of a disruption. Eradication, on the contrary, means that people will not even contract the virus that causes the disease because the virus will not exist anymore. So, as for the case of smallpox, the example that eradication enthusiasts like to bring to the fore since, as stated, it is the only disease that has ever been eradicated, the variola virus that caused smallpox disease in humans cannot be found anymore in nature, and it is kept in laboratories under the auspices of the WHO.13 That being said, we can explore how different the cases of Covid-19 and smallpox are, in order to understand why, contrary to what happened to the latter, the former will become endemic and will not be eradicated. First and foremost, note that when smallpox was targeted for eradication, it was present only in 59 countries (Barrett & Hoel, 2007), not in every corner of earth, as are coronaviruses. As for Covid-19 specifically, when lockdowns were implemented and when vaccination started, Sars-Cov-2 had already spread across the globe and so the rational expectation ought to have been that it will eventually become endemic as all other coronaviruses (Lavine et al., 2021), not that it will be eradicated. The main reason that smallpox was eradicated is that the immunity induced through vaccination lasts for decades while immunity that occurs through infection is lifelong (Taub et al., 2008), and so, after centuries of vaccinations and infections, the WHO announced in 1980 that the disease no longer exists and recommended against repeated vaccinations. On the contrary, the available vaccines against Covid-19 do not provide life-long immunity (not even long-lasting), and the protection they confer wanes after a few months. Most importantly, they do not stop transmission, as we saw. So, leaving aside moral issues, even if we managed to vaccinate the world’s population, the virus would still be transmitted among the vaccinated people and the disease would have been nowhere near to eradication. Even in that case, the result would have been Sars-Cov-2 endemicity through infection and immunization, not eradication. It is also known that natural immunity against Hcovs also wanes over time and that reinfection is indeed possible (Edridge et  al., 2020). However, reinfected persons develop milder symptoms than the ones suffered during their first exposure. Natural  I consider CFR and IFR to be the same here because I discuss only the 95 cases identified among the residents, and I assume that there were no mild or asymptomatic cases that went unrecorded among the remaining 48 residents. 12  In spite of the ineffectiveness of the vaccines to stop transmission, I take for granted that they provide even some short-term protection against severe disease. 13  The remaining stocks of variola virus are used primarily for research purposes (Damon et al., 2014) and also to prevent possible bioterrorist attacks using variola virus (CDC, 2016b). 11

2.5  Wishful Thinking Part 2: Zero-Covid Through Mass Vaccination

25

immunity against Sars-Cov-2 is robust, long-lasting, and stronger than the one elicited by the vaccines (Cohen et al., 2021; Jung et al., 2021; Pape et al., 2021; Turner et al., 2021), but, as is the case with all coronaviruses, reinfection is possible sometime in the future, which makes clear that even if everyone on earth gets infected, the virus will not be eradicated either. As is the case with all Hcovs, when reinfection happens, the agent also develops only mild symptoms. It is noteworthy that until the fall of 2021, an extremely low incidence of reinfections to Sars-Cov-2 (Murchu et al., 2021; Vitale et al., 2021; Abu-Raddad et al., 2021) was reported, and these people developed symptoms that were typically mild to nonexistent, which suggests that naturally acquired immunity confers protection that lasts for years. Similar appears to be the case with Sars-Cov-1 that circulated in Asia in 2003 and eventually became endemic. As we saw above, some people were having robust specific T-Cell immunity even 17 years after the infection (Le Bert et al., 2021). The available evidence on coronaviruses (including Sars-Cov-2) therefore implies that immunity to coronaviruses wanes, which makes reinfection a documented fact and eradication impossible, but is not entirely lost, which guarantees that reinfection causes milder symptoms than the ones caused after the first infection, making it not very much of a public health problem. So reinfection is not a problem for people’s health; it is only a problem for policy-makers if their target is eradication. People will end up controlling Sars-Cov-2 through herd immunity, and some of them will be occasionally reinfected as it happens with most pathogens (Hcovs, influenza, adenoviruses, etc.). Briefly put, reinfection and endemicity put paid to eradication plans. Likening Covid-19 to smallpox makes little sense from a scientific point of view for, apart from reinfection, there are glaring differences between the two diseases that show that trying to find similar patterns between them and, based on these patterns, to conclude that we should eradicate Covid-19, as we did with smallpox creates muddy waters in science. However, we can stress more differences between these diseases to show that the calls for covid-eradication do not put forward a well-­ worked thesis. Smallpox was a terrible disease with a CFR in the region of 30% (Henderson et al., 1999), and I found no IFR estimates so I take for granted that it was the same as the CFR (given the severity of the disease, it seems that there were hardly any smallpox cases not officially recorded, and even if there had been some, it is likely that they were not very common, and so there would hardly be a notable difference between the IFR and the CFR of smallpox, as it happens with coronaviruses and especially with Sars-Cov-2). Moreover, smallpox killed and/or caused serious damage mainly to young people and children. On the contrary, Covid-19 is a threat only to the old people, and its CFR and IFR are, as we saw, very low and nowhere near that of smallpox’s. So, trying to eradicate a disease that kills primarily a particular age-group, namely, septuagenarians and octogenarians, sets a paradoxical precedent in science because these people are vulnerable to a vast array of pathogens all of which share similar features (transmission patterns, IFR, symptoms), which could indeed cause severe damage to their health. All Hucovs, flu-viruses, and bacterial pneumonias are considered to be a threat to those people (Mouton et  al., 2001; Strausbaugh et  al., 2003; Gilca et  al., 2020), and so if eradication

26

2  Viral Mitigation: Weak Theoretical Underpinnings

advocates are to be consistent with themselves, they need to argue in favor of eradication of every possible pathogen that poses a threat to old people. One could argue, of course, that we should eradicate Covid-19 not only to protect the elderly but also to eliminate even the low risk young people face. Again, the same problems appear. For example, H1N1 has an IFR that is in the same region as that of Sars-Cov-2 (0.1% (Fidgerald, 2009) vs. 0.15%, respectively), and so it would make sense to compare these two pathogens. In that case, one should either argue for spending money in an attempt to create vaccines that confer permanent immunity for H1N1 and, in general, for any pathogen with similar features or recognize that eradicating these diseases is an impossible, but not an unpleasant, outcome. Targeting smallpox for eradication was based on the idea that it kills or maims mostly young people and at very high rates. That is definitely not the case with Covid-19 and anyone who argues otherwise either pursues her/his own interests, as politicians did, or is not well informed. Another argument in favor of eradication rests on the idea of “long-covid.” So, some people claim that we should eradicate the disease because despite the fact that it has low death rate, it may have serious implications in the long-run for those infected. No data support this claim, though. Studies have shown that incidences of long-covid are rare and are not of major concern (Sudre et al., 2020). Note further that most symptoms associated with long-covid are often very common in the population, and attributing all of them to Covid-19 is questionable. Several symptoms such as headache, tiredness, anxiety and sleep disorders, which are often associated with long-covid, can have numerous causes and thus, in my view, we should take even these low long-covid rates with a pinch of skepticism for the simple reason that the symptoms in question may have nothing to do with Covid-19. French researchers made exactly this point: we should tread carefully when ascribing symptoms to a disease and argued that physical symptoms after a Covid-19 infection are not to be so easily linked to Sars-Cov-2 for it is possible that symptoms are erroneously attributed to the virus (Matta et al., 2021). This lack of clear correlation between the symptoms and the disease in the long run may lead to quite peculiar results. Data published by the UK government on long-covid showed that children 2–11 years old who had not contracted the virus suffered long-covid at higher rates than the ones who had caught the virus (ONS, 2021b). If we are to take such results to heart, we should conclude that infection with Sars-Cov-2 benefits human health, at least as for these age-groups. The more reasonable conclusion though is that long-covid is not a concerning effect. On the contrary, long-term effects of smallpox included scars on people’s skin and occasionally blindness (CDC, 2016a). We can be confident that these symptoms were induced by smallpox disease since they not very common in the population, and they could hardly be more often found to those that did not get infected with variola virus than to those who suffered smallpox. It seems fair to conclude that comparing smallpox with Covid-19 leads us to the conclusion that Covid-19 is mostly harmless while smallpox was mostly damaging. So the comparison between Covid-19 and smallpox may be useful in order to argue that zero-covid is not only unfeasible but also needless, while in the case of smallpox, it

2.6  The Fundamental Problems of Lockdown Mechanism

27

was not needless and it turned out to be feasible. But drawing upon this comparison to argue in favor of eradicating Covid-19 is totally off the mark. However, let us forget the above for a moment and assume that we manage to eradicate not only Sars-Cov-2 but also all other coronaviruses and all influenza strains. Since coronaviruses, flu-viruses, and similar pathogens are notorious for being transmitted from animals to humans (Tan et al., 2022), a novel coronavirus, or a novel flu-strain, which could likely be as dangerous as Sars-Cov-2 was for old people, could jump from animals to humans, and so, according to eradication advocates, we should eradicate this virus too (it would be inconsistent from their perspective to claim that we could live with such a hypothetical virus for the damage it causes is akin to that of Sars-Cov-2). This could demolish our eradication efforts, and in that case, we should either accept that we have to coexist with these viruses or, else, humanity would be trapped to an endless effort of eradicating diseases both from humans and, in order to be on the safe side and eliminate the risk of animal to human transmission, from animals too. In a phrase, eradicating Covid-19 was never a likely outcome, and those who claimed otherwise (politicians, pundits, et al.) should make a convincing case for eradication based on solid evidence and eloquent scientific rationale. Their claims that this can be done through lockdowns or through mass vaccination are far from being scientifically serious and convincing. And when politicians endorse claims that are not scientifically serious, we should be aware of their motives. Here I focused solely on the scientific implausibility of zero-covid, which, once documented, suggests that the public interest perspective on the issue is flawed. In the following chapters and especially in Chap. 5, I will pay attention to the reasons some politicians adopted a zero-covid policy.

2.6 The Fundamental Problems of Lockdown Mechanism We now know that the postulates upon which the lockdowns were based were mistaken, and we have seen that one of its targets, i.e., to eradicate the disease, was not achieved through lockdowns and is unattainable not only through lockdowns but by any possible means, such as mass vaccination programs. The other target of lockdowns, which was based on the very same postulates that the zero-covid goal has been, was to reduce transmission rates, suppress the disease, prevent hospitals from being overcrowded, and thereby save lives. Before discussing the real-life failures of lockdowns, it is, I think, worthwhile to try to estimate what would have happened in case lockdowns were efficient at reducing transmission and death rates. The standard lockdown strategy includes school-closures, business shutdowns, and curfews. We should also consider tight border controls as a part of lockdown policies. Assume now that country X manages to identify the first cases very early on and imposes a very strict lockdown by implementing all these measures and manages to have very few cases and deaths. After a few months, policy-makers face the dilemma: Are we to lift the lockdown or should we keep it in place indefinitely?

28

2  Viral Mitigation: Weak Theoretical Underpinnings

Note that they know that despite the fact that the mitigation strategy was successful, the virus is still in the community and has not been eradicated. If they lift the measures and people start interacting with each other, then cases and deaths will grow and a new dilemma will appear: Should we implement a second lockdown or should we let the virus run its course? If they choose the second option, then the initial lockdown is as if it was never implemented because it simply delayed the inevitable infections and deaths and did not offer a viable solution. If they choose the first option and they manage to successfully restrict the spread one more time, then the first dilemma bounces back and this vicious circle is likely to continue with no end in sight. Another line of reasoning could be that another country (X’) manages to control its borders prior to the advent of the virus, and so the country is covid-free. Indeed, there is some evidence suggesting that even if full lockdowns do not impact on mortality, there may be some benefit from border controls (Chaudhry et  al., 2020). However, a similar class of problems appears. Borders should be kept closed forever, and no one should enter or leave the country not only in the foreseeable but also in the distant future, given that there will be countries in which the virus will still circulate. But to keep strict border controls in place forever is an unfeasible task. So they will, almost inevitably, become less strict. Once border controls become less tight, the virus will probably enter the country and start to circulate, and so the border-restrictions that were implemented in the first place turn out to be of no avail. The pro-lockdown objection to the above would be that if we manage to halt transmission, then as time passes, we may have available treatments and/or vaccines. Hence, even if the cases increase when the first lockdown is lifted, lives could be saved without a second lockdown since, at this stage, we can implement pharmaceutical interventions. Enforcing a lockdown early on relying on the assumption that effective treatments will be soon available implies that policy-makers have access to reliable sources of information which have reassured them that effective treatments or vaccines would be available for widespread use sooner rather than later. This suggests that treatments or vaccines will be made available within days or a few weeks after the initial lockdown implementation, and so the non-­ pharmaceutical interventions, as are the lockdowns, would be instantly replaced by pharmaceutical ones, namely drugs. If vaccines and drugs do not arrive quickly enough, then the lockdown will either have to be extended for months, which makes it a very costly strategy, or, even worse, to the indefinite future, waiting until treatments are available. Alternatively, it can be lifted, and so cases will spike and thus a second round of lockdown measures would be seriously considered. In the case of Sars-Cov-2, vaccines were available in record time (approximately ten months after the official declaration of the pandemic), and one can hardly claim that they can be rolled out earlier than that in case another novel pathogen starts to circulate. Speaking of the treatments against Covid-19, monoclonal antibodies seem to be life-saving (Dougan et al., 2021), but their effectiveness was vindicated even later than the time vaccines became available which further weakens the pro-­ lockdown case. By and large, there is still no scientific consensus on the preferable drugs, and while this is not the place to elaborate on drug-effectiveness, it suffices

2.6  The Fundamental Problems of Lockdown Mechanism

29

to say that the literature is rather inconclusive on the issue by offering evidence that is hard to reconcile. For example, there is evidence that suggests that anti-viral drugs such as hydroxychloroquine may be promising if used early on (Prodromos & Rumschlag, 2020; Mokhtari et al., 2021) while others suggest that it is not (Axfors et al., 2021). Similarly, some studies indicate that ivermectin is helpful (Caly et al., 2020; Jans & Wagstaff, 2020), whereas others draw the opposite conclusion (Popp et al., 2021). There is also much controversy over whether remdesivir works. The World Health Organization recommends against its use (WHO, 2020c), and there are studies that speak against it (Ader et al., 2022), while the FDA has approved it (FDA, 2020). Whatever the case, it appears to be clear that keeping a lockdown in place until all the scientific uncertainties are resolved is likely to prolong the lockdown for years, which is an unfortunate outcome. For the sake of simplicity though, let us assume that, ideally, policy-makers knew that ten months after the initial outbreak, effective vaccines and drugs will be at our disposal, and let us further assume that we are confident that whenever another pandemic strikes, we will be able to implement effective pharmaceutical interventions within ten months. A ten month time on its own is neither short nor long. It is considered as short or as long within a context. So, the ten month time that took Covid-19 vaccines to become available may be a miraculously short period of time when considered in the framework of vaccine-development (since it typically takes years or even decades) but is a tremendously large period to keep in place lockdowns and similar restrictions. Shutting a country down is costly even if it is done for only one day, and the more such policies are in place, the higher the cost. Prior to 2020, this was common place among the scientific community. Large-scale quarantines were regarded as having so extreme negative consequences that should not be seriously considered under any possible circumstances (Inglesby et al., 2006). Policy-makers and lockdown advocates have failed to offer a thorough cost/benefit analysis that would suggest otherwise, which is what they ought to have done prior to lockdown-implementation since it was clear that, according to our state of knowledge until 2020, such a disruption of social and economic life is damaging, and thus those who argue in favor of lockdowns bear the burden of proving that the cost is worth suffering it. So even if we were sure that lockdowns reduce transmission rates and that in about ten months effective vaccines and drugs will be rolled out, no theoretical or empirical account indicated that even short-term lockdowns are worthwhile, let alone extended lockdowns that last for ten months. But there are moreover problems regarding the need for a lockdown even if vaccines or drugs arrive immediately, i.e., within days after a novel pathogen appears. Assume for example, that a vaccine or a drug is available within a week or two after the outbreak. In that case, a lockdown strategy, apart from being costly, it becomes moreover needless since people will immediately benefit from the pharmaceutical interventions and there is little point to try to stop the spread. The bottom line is that even if we assume that lockdown strategies work as their proponents thought they would, they offer only temporary results by simply delaying the inevitable infections and deaths; they do not address the problem and are an unbearable strategy to be kept in place for months, let alone in perpetuity. Even if

30

2  Viral Mitigation: Weak Theoretical Underpinnings

drugs arrive quickly, that would not be so fast in order to evade lockdown-harms, and even if they do arrive immediately, then there is no need for a lockdown. So even if all people faced pretty much the same risk of dying, even if asymptomatic transmission was common, and even if lockdowns remarkably reduced the rate of infections and deaths, the proper public policy approach, since the virus is not eradicable, would be to identify those most at risk and try to protect only those people. Therefore, since even an idealized version of lockdowns simply prolongs the problem instead of tackling it, public policy decision-making should refrain from implementing it. In the real world though, things are even worse for lockdowns.

2.7 Lockdown Failure Lockdowns are not a mechanism for disease-eradication, and the virus still circulates even if the society is locked down. Therefore, a serious methodological problem shows up. In an attempt to protect everybody, lockdowns make each one equally exposable to the virus. So, since the same rules apply to everyone, younger people, who face almost zero risk from the virus, have equal chances to contract the virus as those over 70, who face high likelihood of death. To reduce the likelihood of the elderly to catch the virus, the youngsters should be exposed to the virus and the elderly should take special protection. Consider the following scenario: person A is 20  years old, and person B is 85 years old. Both are enforced to stay at home and leave it only to go to the grocery store or to have a walk nearby their households. Clearly, both take the same risks and so both have the same probability to be exposed to the virus, despite the fact that for person B, the IFR is very high (well above 1%), and for person A, it is very low (near to 0%). In this respect, lockdowns, instead of protecting them, in fact expose septuagenarians and octogenarians to the virus because if these people are to be protected, they should take fewer risks than the younger population and vice versa: the younger people should go on with their lives so that the virus would circulate mostly among them by leading to the development of robust population immunity and thus making it more difficult to infect the elderly. Naturally acquired immunity against Sars-Cov-2 is strong, long-lasting, and protects against serious cases of reinfection (Cohen et al.; 2021; Jung et al., 2021; Pape et al., 2021; Turner et al., 2021), and so if people that have a near to zero risk of dying live their normal lives and one way or another contract the virus, they will typically recover quickly, and they will gradually build up population immunity, which will protect the vulnerable. But if the vulnerable and the youngsters are equally exposed, then the former will be more easily infected and since they face much higher risk of dying, the death rates will probably increase and the levels of population immunity are likely to be lower. The more low-to-zero-risk groups contract the virus, the more the high-risk groups are protected and conversely, the more low-to-zero-risk groups are protected the higher the chances of the elderly to get infected. So, Person B would have been safer if individuals under 70 were allowed to live their lives, and the government

2.7  Lockdown Failure

31

implemented some targeted measures for B’s age-group so that people at their age would avoid—at least for a period of time—some risky activities (such as going to the grocery store).14 Thus lockdowns do not provide special protection for those at risk which is what the data imply that should have been the case (Karadimas, 2022b). The counterargument to this is that by leaving young people go about their lives will result in older people getting more easily infected because cases will increase and young people will transmit the virus to older ones, especially if they share a common household. However, the available theoretical knowledge and sound empirical data from several countries across the globe suggest that if no lockdowns are enforced, the chances of the elderly to get infected are not increased, and we cannot rule out the possibility that they are in fact reduced. Firstly, the objection seems to hinge upon the claim that in the absence of lockdowns, people do not take by themselves commonsensical preemptive measures. That is not true and it seems reasonable to expect that the contrary is in fact the case. When older people know that a virus that poses high risk to them circulates, they would voluntarily adjust their behavior, at least for some time, in order to protect themselves. In the early stages of mass hysteria, most people would follow the emergent norms, but if no lockdowns are imposed and a data-driven public health message is conveyed, then mass hysteria will gradually weaken and young people could live normally, while the older would be more cautious. Secondly, policy-­ makers failed to take into account that household transmission is the main setting of transmission (Luo et al., 2020), which should have made it clear that the more all the members of a family stay at home, the more likely for them to get infected, lest one member of the family develops symptoms. Therefore, the likelihood of older people to get infected if they live with youngsters increases not if life goes on as normally as possible (which typically includes that most members of the family spend many hours of the day out of home) but under lockdown since all members of the family are stuck at home and thus household transmission is likely to thrive. Thirdly, we know that a particular group of young people, such as schoolchildren, do not cause infections to spiral out of control (Munro & Faust, 2020; Rotevatn

 Of course, it is not entirely accurate to claim that all people were equally exposed to the virus under lockdown, for the so-called “essential workers,” such as workers in a grocery store, were exposed to the virus much more often than other people during lockdowns. But since they are typically under 70 years old, they faced little to no risk from the virus. However, if we follow the prolockdown argumentation, these people need to be protected, too. By letting them go to work, it is as if supporters of lockdowns admit that they expose certain people to a serious death-threat while they protect others—usually white-collar workers who can work from home. And even though from an anti-lockdown perspective we should not bother that much about possible virus-related health problems that essential workers could face, from a pro-lockdown viewpoint, this is a glaring inconsistency. Either all people face serious risk and all should be protected, or we should protect only those at really high risk. But to expose certain groups of blue-collar workers to the virus while protecting the elites and simultaneously claiming that the virus is a real threat to everyone is not a coherent strategy. 14

32

2  Viral Mitigation: Weak Theoretical Underpinnings

et al., 2022),15 which implies that closing schools did not reduce the risk of infection for the vulnerable population. Fourthly, we saw that asymptomatic transmission almost never happens, and it is therefore simply mistaken to claim that every single person is likely to spread the disease and that each individual would, unbeknownst to him, pass it on to a vulnerable person. Thus, by quarantining those without symptoms, i.e., the healthy ones, transmission is not reduced. Recall moreover that, as already stated, the WHO suggested, at least until 2019, that in order to mitigate respiratory diseases, only those with symptoms are to be isolated, not those that were exposed to an infected person (WHO, 2019), let alone those that did not even come into close contact with symptomatic individuals. Thus, quarantining everyone is a strategy aberration with unclear, to say the least, benefits. Even fomite transmission has been seriously questioned since the detected viral fragments on fomites are dead and are not able to cause infections (Mondeli et al., 2020; Onakpoya et al., 2021) which casts further doubt on the pro-lockdown worldview of an omnipresent virus that can be transmitted in every conceivable way. So, since household transmission drives the pandemic, since closing schools and restricting the movement of healthy people does not reduce infections, and since fomite transmission is yet to be proved, there is no theoretical reason to believe that if no lockdowns are in place, transmission rates would increase dramatically, and older people would face a higher risk of being infected than they do under lockdowns, where they are much more exposed to the possibility of household transmission. Data from countries that locked down and countries that did not provide moreover strong empirical reasons to reject the claim that lockdowns reduce transmission and deaths by verifying that the viral trajectory is irrespective of non-pharmaceutical interventions. Indeed, several studies found that full lockdowns and other strict mitigation measures are not associated with mortality (Larochelambert et al., 2020; Herby et al., 2022) and that less severe interventions lead to pretty much the same results (Haug et al., 2020). Researchers compared data from countries like Sweden and South Korea that did not shut everything down with countries that have had a very strict lockdown, which included, among others, countries like the UK, Belgium, and France, and concluded that no significant differences in case growth and decline were identified (Bendavid et al., 2021a), which puts paid to lockdown strategy since the purpose of its imposition was exactly to stop the virus and to save lives. Lockdowns therefore, appear to have fallen short of achieving their main goal which was to stem remarkably the spread of the virus and to reduce the virus-related mortality. For example, Belgium, Italy, the UK, Peru, and France all had a very strict lockdown, and all suffered a higher death toll (per million of the population) than Sweden which had a no-lockdown strategy (OWD, 2022a). Moreover, Sweden is having lower death rates than the average rates in the EU (OWD, 2022b). But for Sweden, all members of the EU locked down. This seems to fit the description of

 Pre-existing immunity from other coronaviruses as well as the reduced ACE 2 receptors in children appear to explain the very low infectivity in these age-groups (Steinman et al., 2020). 15

2.7  Lockdown Failure

33

lockdowns as a setting under which all age-groups are equally exposed to the virus, and thus death toll increases under lockdown. The cases may follow a similar pathway (in the sense that the epidemic wave peaks for two weeks or so and then it goes away irrespective of the non-pharmaceutical interventions that are implemented), but in no lockdown settings (such as in Sweden), it is likely that younger people are exposed much more than the elderly, and so the number of deaths turns out to be lower since the subpopulation that is at high risk of death is, either intentionally or inadvertently, protected because the virus circulates mainly among the low-to-zero-­ risk groups and strong, long-lasting natural immunity is developed, something that is not happening when lockdowns are implemented and all follow the same rules, such as in Belgium, Italy, the UK, Peru, and France. Greece is another example that shows clearly that the increase and decline in cases is irrelevant to the measures that are in place and that the death rates may be higher than otherwise if lockdowns are in place. Greece was on a very strict lockdown from early November 2020 to early May 2021. Over that period of time, cases spiked twice and the deaths were higher than the death rates in Sweden where life went on pretty much normally (Greece’s total death toll is also higher than Sweden’s) (OWD, 2021f). It becomes overwhelmingly clear thus that since the virus circulates across the globe from January 2020, if not earlier than that (Tichopád et al., 2021; Paixao et al., 2022), and since repeated strict lockdowns were in place in most countries in the EU and in the UK from March 2020 to May 2021, if lockdowns were the game-changer their advocates thought they would be, then Sweden ought to have substantially higher covid-­ related death rates than countries with lockdowns, but, as it is by now clear, that did not happen. In the USA, South Dakota makes another clear case against lockdowns. In March 2020, all States locked down, with the exception of South Dakota. While the governor closed schools from March 2020 to May 2020 (Ballotpedia, 2021), no business closures and no curfews were in place. Schools reopened in May 2020, and since then there have been literally no restrictions in South Dakota and life has pretty much been “business as usual.” The death rates in South Dakota are lower than those in many states with harsh repeated lockdowns (per 100,000 population) (Statista, 2021), which one more time suggests that lockdowns either have no impact on mortality at all or, even worse, they exacerbate things. The discussion vis-à-vis the transmission dynamics as well as the inability of lockdowns to significantly reduce cases and save lives, appears to refute another strand of pro-lockdown argumentation according to which lockdowns prevent hospitals from being overwhelmed and so optimal treatment for all patients is possible. Hospitals during winter time are typically pressed, and it appears that during the Covid-19 pandemic, they did not face substantially higher pressure. Moreover, since it has been shown that lockdowns do not impact on the viral trajectory, there is no sound reason to assume that hospitals would be overflown with covid patients if no lockdowns are in place. The assumption of overwhelmed hospitals in no lockdown settings is based on the mistaken conviction that under no lockdown, cases will be dramatically higher than are under lockdown, and thus a remarkably higher number of severe cases would ask for urgent medical treatment. Since cases follow pretty much the same pattern with or without a lockdown, so do hospital

34

2  Viral Mitigation: Weak Theoretical Underpinnings

admissions. On the contrary, since, as we saw, many countries that imposed lockdowns suffered higher death toll than no lockdown countries, and since all agegroups take the same risks in lockdown settings, which means that the older people are not protected, it then appears not to be an implausibility that hospitals may be more pressed under lockdowns than they are under no lockdowns. To put the matter in a nutshell, the more the elderly are exposed, the higher the likelihood of hospitals to be pressed, and the more the younger people are exposed and built up population immunity, the more the elderly are protected and the less the hospitals are pressed. The CDC has released data suggesting that hospitalization rates for Covid-19 are in the region of hospitalization rates for influenza (CDC, 2020b, 2022a), and we have no data which could imply that mitigation measures impact on hospitalization rates and that hospitals in Sweden and South Dakota were more pressed than hospitals in lockdown regions. Those who claim otherwise bear the burden of proving their claims, but at least so far, they have not done so.

2.8 Pre-existing Immunity vs. Shutdowns A key figure that determines Covid-19 mortality and further undercuts lockdowns and other mitigation measures is pre-existing immunity. The levels of pre-existing immunity in the societies seem to vary from 20% to 50% (Doshi, 2020a; Mateus et al., 2020), and one can hardly oppose the claim that countries with higher levels of pre-existing immunity suffer fewer deaths than countries with lower ones. It is commonly estimated for example that Asian countries faced low death-rates from Covid-19 due to the high levels of pre-existing immunity in these populations (Braun et al., 2020; Bolourian & Mojtahedi, 2021) and not due to the strict lockdowns, which are commonly praised by lockdown advocates and the press (Harrison & Kuo, 2020). If that was due to lockdowns, then all countries with tight lockdown would have the same death-rate as the Asian countries, but, as we saw, that did not happen. Another possible explanation for this is the lower average longevity in these countries which implies that the population in these countries is mainly young and remains largely unaffected by Sars-Cov-2 because people die prematurely due to other causes. If the few deaths in Asia are to be attributed to the lower average lifespan that is a matter of fact in some Asian countries, such as in Afghanistan, Syria, or Yemen (World Atlas, 2019), then Asian countries with low death toll should have average longevity significantly lower than that of Western countries. However, many Asian countries with very few covid-related deaths have pretty much the same life expectancy as the one in the EU and in the USA. For example, Japan, Singapore, and South Korea all have the same life expectancy as most countries in the West (84.2, 82.9, and 82.7, respectively) (World Atlas, 2019), albeit they all suffered much fewer deaths from Covid-19 (per million population) than the EU and the USA (OWD, 2021b). Consider further that, as we saw, South Korea and Sweden are among the countries that did not impose a lockdown and suffered lower death-toll in comparison to most countries that did enforce lockdowns. But among these two

2.9  Modeling Drawbacks: Poor Inputs, Poor Outputs

35

anti-lockdown countries, it is South Korea that has the fewer deaths (per million population) (OWD, 2021a). It seems compatible with the evidence to claim that the levels of pre-existing immunity have been the game-changer in all these cases. In general, shutting down a country with high levels of pre-existing immunity is pointless because herd immunity is almost built up even prior to the advent of the virus. Shutting down countries with low levels of pre-existing immunity is useless because, as we saw, mortality is not reduced, and it is possible that virus-related deaths may sky-rocket. It seems fair to conclude that lockdowns did not help us address the pandemic in any meaningful way.

2.9 Modeling Drawbacks: Poor Inputs, Poor Outputs While there are practically no empirical data in favor of lockdowns, advocates of non-pharmaceutical interventions bore to the fore some models that claimed that lockdowns and other mitigation strategies have saved many lives. They go so far as to say that lockdowns and other mitigation measures, such as (mandatory) mask-­ wearing, have saved thousands or even millions of lives. The most commonly discussed and cited model is the one by Flaxman et al. (2020), which appeared a few months later in 2020 after the appearance of the already mentioned paper presented by modelers at the Imperial College in March 2020 (Ferguson et al., 2020), which included the overestimate of an IFR and as we saw took it to be in the region of 0.9% and which was eventually published in the journal Nature in June 2020. These models enthuse over NPI and conclude that especially lockdowns are effective in halting the spread of the virus and in saving lives. The other model I would like to discuss is that of (Chernozhukov et  al., 2021) which appeared in the Journal of Econometrics in January 2021, though it was available online by 17 October 2020, and which also argues in favor of NPI and it emphasizes that mask-wearing, in particular, could reduce transmission and save many lives. The Imperial modeling which is famous for guiding the world’s response to the virus, predicted in March 2020 that if no lockdowns are imposed during the first pandemic wave, the UK will suffer 510,000 deaths and the USA will suffer 2.2 million deaths, while the actual death rate in both countries amounted to only a tiny fraction of that. To give the reader a glimpse of its failure, modelers from Uppsala University projected these estimates to Sweden and argued that in the absence of a lockdown, Sweden would face from 52,000 to 183,000 deaths with a median estimate in the region of 96,000 deaths till July 1 2020 (Gardner et  al., 2020).16 By  This is a pre-print publication. Later in 2020, two of the coauthors of this study published a peerreviewed piece (Kamerlin & Kasson, 2020) in which they admitted that they overestimated the predicted death toll, but they put their overestimate down to the voluntary changes in behavior that took place in mid-March 2020 in Sweden. However, it appears that voluntary changes in behavior by the general public cannot explain such a difference in the death-rates for they are akin to light lockdown-like interventions which, as the data suggest and as I argued in the previous section, 16

36

2  Viral Mitigation: Weak Theoretical Underpinnings

December 2021, Sweden had just above 15,000 deaths, and until August 2022 there are 19,810 reported coronavirus-related deaths (FOHM). The later version of the Ferguson et al. (2020) model, which is Flaxman et al.’s (2020), tried to estimate how effective NPI were during the first wave of the pandemic. The main line of criticism against this and similar models that are overly confident that mitigation measures work is that people adjust their behavior even without mitigation measures, and thus lockdown-style measures become redundant. I disagree with this criticism, for while it rejects lockdowns, it seems that it accepts that social-distancing (either mandatory or voluntary) saves lives. It appears that it is not the case, since, as we saw above, when healthy people restrict their moves, the impact on transmission rates is zero, exactly because asymptomatic transmission is a fallacy. Moreover, when young people practice social-distancing, this makes it easier for the virus to infect the vulnerable groups, and so it is hardly the case that voluntary social distancing reduces death rates. Self-isolation makes sense when only infected people (and perhaps the vulnerable) are isolated. When asserting that voluntary social distancing by the general population is effective, it seems to me as if one argues in favor of the equal infectivity thesis which assumes that healthy individuals infect each other, which is mistaken. A second and perhaps stronger point often made against model-based claims that lockdowns work is that lockdowns often come too late, when the wave has peaked and cases have already started to decline and thus there is no strong causal relation between the decline in cases and the lockdown (Wood, 2020). Indeed, the fact that cases fall when an NPI is in place does not imply a causal connection between the two, for as we have seen, the case growth and decline appear to be irrespective of the measures that are in place. A belatedly imposed lockdown for example, seems to have happened in Germany in March 2020 whereby the cases had started to fall prior to the implementation of the lockdown (Kuhbandner et al., 2022). However, even this objection to lockdowns seems to allow for the possibility that lockdowns are grounded in science and that the only problem is that they may be belatedly imposed which is, as we have seen here, not accurate presumption. Therefore, I do not fully endorse none of these two claims for if we assume that lockdowns halt transmission rates if imposed in time or if a voluntary lockdown is also able to save lives, then modelers could have been right into claiming that “fewer interactions lead to fewer deaths.” But the data we examined in the previous sections suggest that lockdowns do not impact on transmission rates no matter how early they are imposed and how strict they are and as we will see in a while there was never sound basis to claim that lockdowns can do so. A third way to

simply guide the epidemic wave in other settings and they do not reduce transmission and death rates. Moreover, the voluntary self-isolation was not prevalent for a very long time in Sweden, and people returned to their normal lives by mid-April 2020 and lived normally throughout the winter 20–21. To the best of my knowledge, the authors have not addressed this issue, and they have not explained why in the absence of a lockdown and with people living normally in Sweden for the best part of the pandemic, Sweden not only did not face 96,000 deaths but also faced fewer deaths per million of the population than most countries with strict lockdowns. So their conclusions remain untrustworthy.

2.9  Modeling Drawbacks: Poor Inputs, Poor Outputs

37

assess the credibility of an epidemiological model, and one which often encompasses the above two comments, is through what experts call “model comparison.” The Imperial College team has developed a similar model to that of Flaxman et al. (2020), which focuses on the USA (Unwin et al., 2020) and similarly overemphasizes the effect NPIs have. Scholars have compared these two models and a third one, which is a hybrid version of the two, and suggest that the Imperial modeling yields results that are not robust to reasonable changes in model-specification and that the inferences it draws collapse upon closer inspection (Chin et al., 2021). While this kind of modeling has already been extensively criticized and its validity has been tarnished (Ioannidis et al., 2020), a few more comments are worthwhile since there are some serious problems with the way pandemic modeling is constructed and evading them in future would be helpful when predicting pandemic scenarios. This is important because weak pandemic modeling of this sort has some history in epidemiology. For example, Ferguson et  al. (2005, 2006) made highly questionable estimates on the avian flu (H5N1) and, in general, on influenza strains by holding, contrary to empirical knowledge, that containment measures are effective in stopping them. Models are, in a similar way to thought experiments, hypothetical constructions in which empirical postulates are embedded (Karadimas, 2022a) and then several calculations and perhaps ad hoc hypotheses follow. Thus, the conclusions that are drawn are highly dependent on the empirical part of the model. If this is robust and well-documented, then the result will likely be so. If not, the result is likely to be misleading or even off the mark. If, on the contrary, there are no empirical statements and the model is merely counterfactual, then its assumptions are to be taken with a pinch of salt, though, of course, not a priori rejected. Since the later publication by the Imperial team, has been criticized in the literature along the lines I just discussed and since its estimates on the efficacy of NPI have been rejected by the data, I will focus mainly on the initial paper and on methodological problems related to it, since it is this paper that was exploited by politicians to justify lockdowns and the one that lockdown advocates brandished back in March 2020 to claim that lockdowns is the only way to avoid a health catastrophe. The structure of the model offered by the Imperial College is simple. It is based on the standard epidemiological SIR method (S  =  Susceptible, I =  Infected, R = Recovered), and it starts with the assumption that the more people interact, the higher the transmission rates and thus the higher the death rates, and therefore, the less people interact, the lower the transmissibility and the lower the death rates. In my view, the first problem is that it seems to postulate what is to be proven. Authors seem to take for granted that sweeping mitigation measures work and that more measures are taken the higher the efficiency of the mitigation efforts and then they go on to prove this claim which seems to be a circular way of reasoning. Of course, the authors could reply that the reasoning is not circular and that the premise that “the less people interact, the fewer the death rates,” is an empirical condition and that they simply try to apply it to the case of Sars-Cov-2. But, as we saw, our knowledge until early 2020 indicated a paucity of data to back up such claims, especially when a respiratory disease is taken into account. While never in the past a global lockdown was in place, some mitigation measures had been from time to time

38

2  Viral Mitigation: Weak Theoretical Underpinnings

implemented, and there was general agreement that they are pointless from a scientific point of view. For example, the UK Government recognized that “It will not be possible to halt the spread of a new pandemic influenza virus, and it would be a waste of public health resources and capacity to attempt to do so” (UK Gov., 2011, 28) while during the Asian Influenza Pandemic in 1957–1958, public health experts rejected lockdown-like interventions such as school closures and prohibition of public gatherings, because they claimed, quite astutely, that even if these measures are in place, people would catch the virus elsewhere (Henderson et al., 2009, 270). The historical example of the 1918 Great Influenza Pandemic offers a pre-2020 strong empirical case against NPI. The 1918–1920 pandemic was a terrible one and resulted in about 40 million deaths or, equally, it killed about 2.1% of the then world’s population. To have a clue of how harmless Covid-19 is in comparison to that pandemic, the 2.1% of the world’s population in 1918–1920 equals 150 million deaths adjusted to the world’s population in 2020 (Barro et al., 2020), and as we already know the death toll from Covid-19 is nowhere near to that number. Moving on to the NPI that were implemented at some places, especially in the USA, during that period of time, these included school closures, prohibition of public gatherings, and quarantine of the infected and of the people that came into close contact with an infected person. Epidemiologists demonstrate that their effect in reducing death rates was near to zero, though they may have delayed some of them for a while (Markel et al., 2007; Barro, 2020). Based on the precedent of the 1957–58 and the 1918–20 pandemics, therefore, it is highly problematic to posit that “the more mitigation measures are taken, the lower the incidence of the disease,” as modelers from the Imperial College did, for it is simply not a well-grounded empirical statement, and relying on such dubious assumptions may lead you to totally mistaken predictions, as it eventually happened with the Imperial modeling. There are further drawbacks with the premises that were used in this model. As stated, they take as median IFR an estimate far higher than the actual one, which also played a part in their extremely high death-estimates. To be fair with the authors, they recognize that the virus does not kill at the same rate across all ages, but their calculations apply a median IFR 0.9% to the general population which grossly exaggerates the expected death toll. Second, while they recognize that symptomatic people are more infectious, they assume that asymptomatic transmission is rather common (Ferguson et al., 2020, 4), which is also a wrong premise. Third, they make no reference to the possibility of pre-existing immunity, a scenario totally absent from their estimates. Given the recent H1N1 example, against which a sizeable chunk of the population had pre-existing immunity, to ignore such a possibility when a new coronavirus appears is a mistake. So they appear to hold that all people die at high rates, that even healthy people can transmit the virus, and that all are susceptible to the disease. It thus comes as no surprise that they ended up with such overestimates. The assumptions upon which modeling was based is obviously a problem made also by Flaxman et al. (2020). They take for granted that NPI decrease transmission rates, and they assign a decline in cases to lockdowns despite the fact that it was pretty clear before 2020 that NPI hardly work and became even clearer by May

2.9  Modeling Drawbacks: Poor Inputs, Poor Outputs

39

2020, that in no-lockdown countries cases not only do not increase indefinitely,17 but they peak and fall in the same way this happens in locked down countries, which indicates that there must be another explanation for the decline in cases. What explains the peak and troughs in cases is still not clear and pre-existing immunity or the development of immunity certainly plays a role in that, but what seems to be rather clear is that lockdowns do not affect transmission. Authors also fail to make age-adjusted estimates of the risk of death, which is also unfortunate when estimating death-tolls. They also seem to apply the equal susceptibility thesis, for they consider only antibodies as an indication of the prevalence of immunity in the population, and so they concluded that in summer 2020, very few people were insusceptible to the disease, which does not follow by the data and also distorts the predictions. Sweden, South Korea and South Dakota are the live examples that refute the nearly all the predictions made in these models. The second model I mentioned, (Chernozhukov et al., 2021), also champions the effect of NPI, and they also make the same methodological mistakes as almost all pandemic modeling. They take at face value that all types of NPI work, which is not a matter of fact, and they calculate how many lives they saved while they were in place and how many they were lost in the absence of such mitigation measures. In other words, their estimates rely on non-robust empirical statements. In particular, the NPI they take into account are almost all the contents of a full lockdown strategy: mandatory face masks (they consider them especially for employees in public businesses), stay-at-home orders, closure of K-12 schools, closure of restaurants except take out, closure of movie theaters, and closure of non-essential businesses. This affects the causal inferences they draw and leads them to make spurious correlations between observed events. They assign throughout the paper an observed decrease in cases to an NPI that was in place at the time cases started to fall, but since science does not give us strong empirical evidence to claim that such NPIs reduce transmission, it is not a well-worked inference to claim that a decline in cases is due to NPIs. Since we have already discussed the failure of lockdowns and lockdown-like interventions to halt the spread of the disease, it is important to pay attention to another type of NPI, and to its supposed efficacy, that of face masks, the efficiency of which the authors stress in the abstract. The way the authors treat the issue of mask wearing reveals that they make dubious postulates and that they draw almost arbitrary causal inferences. Prior to the Covid-19 pandemic and throughout this crisis, the evidence on the efficacy of face masks has been very scant, and in fact, most of the available ones indicated that masks do not stop the spread of a disease. This was actually known even prior to the advent of Sars-Cov-2 since masks had failed to help reduce the transmission rates of influenza strains. There are three categories of the masks used: cloth masks, medical masks, and N95 masks. Several studies have compared which one of them is more efficient and whether the use of

 Another model published by Brauner et al. (2021) also arbitrarily correlates a decline in cases with the policies that are in place. 17

40

2  Viral Mitigation: Weak Theoretical Underpinnings

each one of them by the public could help curb the spread of a respiratory disease. It is generally accepted that the gold standard for making such comparisons and in general to have a reliable empirical estimate on several issues is randomized control trials (RCT) (Campbell & Stanley, 1963). Researchers have compared cloth masks, medical masks, and a control group to examine whether they impact on influenza transmission. The control group in this study comprised mask use in a high proportion of participants, and thus there was no maskless group. The highest infection rates were reported among those who wore cloth masks. The authors concluded that cloth masks are inefficient and should not be worn by the general population (McIntyre et al., 2015). The results of this RCT indicate that surgical masks could be more efficient than cloth masks, but since there was not a “no mask” control group the study remains agnostic on whether surgical masks are superior to no mask. The CDC discusses 10 RCT on face masks that appeared in the literature from 1946 to 2018 and did not find significant reduction in influenza transmission either when cloth masks are used or when medical/surgical masks are used and similar results became available from studies that examined the use of face masks in university and household settings (CDC, 2020a).18 Another RCT compared medical masks to N95 respirators and concluded that the infection rates were at about the same region (Radonovich et al., 2019) which indicates that neither N95 nor medical masks are effective. A Cochrane analysis of the literature on masks prior to the Covid-19 pandemic arrives at a similar conclusion: No significant difference in infections between masked and unmasked groups (Jefferson et al., 2020c). Thus, when Sars-Cov-2 arrived, it was quite clear that none of the available masks were efficient at reducing the transmission of influenza and that mask-wearing by the general population does not impact on transmission. The science on face masks did not change since then, as politicians and lockdowners often contend. A top-notch RCT, which focused on Sars-Cov-2 transmission, was published in November 2020 and showed no statistically significant differences in the infection rates between the group that wore face masks and the no-mask group. The between-group difference was −0.3 percentage point (95% CI, −1.2 to 0.4 percentage point; P = 0.38) (odds ratio, 0.82 [CI, 0.54 to 1.23]; P = 0.33) which is negligible (Bungaard et al., 2021).19 Another RCT which appeared during the pandemic conducted in Bangladesh and also suggests either limited or no efficacy of masks (Abaluck et  al., 2021). Similarly, in February 2021, the ECDC reported that the evidence on the efficacy of face masks is scarce and that further research is needed (ECDC, 2021). Since the model by Chernozhukov et al. (2021) was published in print in January 2021, and its online version appeared in October 2020, it can be said that they were unaware of the Bundgaard et al. (2021) RCT and of the ECDC report, but the same cannot be said for the several studies that were published before the pandemic, all of which implied that masks are useless.

 Another meta-analysis which also reviewed all the pre-pandemic studies on the efficacy of masks also concluded that masks are ineffective (Liu et al., 2021). 19  The study was updated since then, but the results remained the same. 18

2.9  Modeling Drawbacks: Poor Inputs, Poor Outputs

41

Surprisingly enough, the authors seem to have been aware that no RCT shows that masks reduce transmission (Chernozhukov et al., 2021, 26), and nonetheless they posit that they are effective in stopping the spread! In support of their claim, they cite some studies which make things even worse. One of them is a pre-print which declares that it reviews the literature on masks, though it mistakenly overemphasizes the results of modeling on masks and downplays the results RCT bear on masks, and hence it ends up with non-robust conclusions (Howard et al., 2020).20 Another pre-print study they cite has perhaps been retracted, for it could not be found online21 and two peer-reviewed studies that discuss issues totally irrelevant to masks and citing them to enhance the claim that masks work is not a fruitful strategy. The first one, by Hou et al. (2020), examines in detail how Sars-Cov-2 infects the upper and the lower respiratory tract and the other explores viral shedding in 94 confirmed Covid-19 patients (He et al., 2020). The word “mask” is simply uttered one time throughout each paper. None of these studies makes a case in favor of masks and neither against mask-wearing for they simply deal with other subject matters. So, to ignore all RCTs that show that masks do not work and to postulate that masks reduce cases based either on studies irrelevant to masks or on theoretical constructs with no empirical grounding and then to draw conclusions by taking this premise as gospel should make us incredulous on the inputs used in the model. And insofar as we are skeptical about the inputs, we will be skeptical about the outputs as well. The authors claim that their counterfactual experiment indicates that mandating face masks for public employees early in the pandemic “could have reduced the weekly growth rate of cases and deaths by more than 10 percentage points in late April and could have led to as much as 19 to 47 percent less deaths nationally by the end of May, which roughly translates into 19 to 47 thousand saved lives” (Chernozhukov et al., 2021). First and foremost, why only on public employees and not on private employees or on everyone? Are employees in the public sector more infectious than others? No data imply so and the RCTs suggest that even if the entire population wore masks the virus would still circulate pretty much as if no one wore them, as it indeed happened in several countries in which mandatory mask-wearing for everyone was in place (Guerra & Guerra, 2021) and we have no good reasons to argue that if public employees wore them transmission rates would be lower. Moreover, the way the authors relate mask-wearing with reduced cases is a classic example of flawed causal inference. To put it a bit bluntly, the fact that a storm and a barometer falling occur at the same time does not suggest that the storm was caused by the barometer falling or that it is explained by it.22 More generally thus, the fact that two events or phenomena are observed at the same time does not on its own imply that there is a causal connection between the two. If we had robust empirical findings that indicated that masks work, then such a correlation could  This study later passed the peer-review muster (Howard et al., 2021), though the methodological problems remain. 21  https://www.medrxiv.org/content/early/2020/08/04/2020.07 22  This example is often used by philosophers of science when discussing the intricacies related to causality (van Fraassen, 1980, 104–105). 20

42

2  Viral Mitigation: Weak Theoretical Underpinnings

have been more sound, but since there are no such data, assigning reduced transmission rates to mask wearing (by public employees in particular!) and making estimates on reduced death-rates based on the very same assumption, is equal to assigning these results to, say, the increased levels of cappuccino consumption in the population while cases fell. Both claims have very little scientific validity. Similar types of problems appear to most pandemic modeling which includes not only models like the ones discussed that try to claim that NPI saved lives but moreover with another model that attempted to discredit the “Focused Protection” strategy which was the proposal of the Great Barrington Declaration (or “shielding strategy,” as it is sometimes called) and thus indicate that lockdowns and masks are the scientifically informed policy, for their conclusion is that if focused protection was opted instead of lockdowns, then we could have suffered tens of thousands of deaths more than the ones actually occurred (Smith et al., 2022). This conclusion contrasts what happened in reality. Since this paper was published in April 2022, the fact that it ignores evidence from Sweden and South Dakota which, as we saw, had lower death toll than most pro-lockdown regions in the EU and the USA respectively, is bewildering and could be on its own enough to make us reject the conclusions of this construction. However, it is worthwhile to focus on the underlying assumptions the authors make which appear to render the model inherently problematic for they too are based on assumptions with little empirical strength. While the authors take into account that the IFR differs sharply between the old and the young and they moreover seem not to posit that there is asymptomatic transmission—both constitute a substantial improvement over the previous models and especially with respect to the Imperial modeling—the other assumptions they make also lack robust empirical grounding, and thus, in the same mode to other models, they are prone to similar sort of criticism. First, as just mentioned, the model ignores real-world data which is unfortunate and a serious drawback. Instead of discussing how the virus circulated in no lockdowns countries and addressing the very issue all lockdowners and all public-interest theorists need to tackle, which is to explain why Sweden and South Dakota performed better than many countries with strict measures, the authors abstain from data that could be elicited from real world and call us to imagine a (non-existent) city with 1,000,000 population which they divide into three groups: a high-risk group that lives in the community, a high risk group that lives in care home facilities, and a low risk group. Another notable drawback of their model is that they take them all to be equally susceptible, which contradicts core scientific findings, such as pre-existing immunity in the population, and thus while the authors do not share the equal vulnerability and the equal infectivity thesis, they seem to tacitly posit the equal susceptibility thesis, which is not a well-grounded assumption, and altering it and assuming that a chunk of the population is already immune would have significantly impacted on their results. Their construction also hinges upon the SIR model, and as all models did, Smith et al. heavily rely on the premise that the less people interact, the lower the transmission rates and the lower the death rates, which is, as repeatedly emphasized, a not well-documented hypothesis for it does not take onto account theoretical and empirical knowledge about transmission mechanisms. This reveals an internal

2.9  Modeling Drawbacks: Poor Inputs, Poor Outputs

43

inconsistency in their argument; the claim that lockdowns reduce transmission rates was initially based on the mistaken assumption that asymptomatic transmission is common. Since the authors are aware that only infected people are able to transmit the virus, then the claim that “the less people interact the lower the transmission rates” is becoming even more arbitrary that it already was. Indeed, it appears that their claim that if it is only the vulnerable that are shielded, then transmission rates will be so high that the vulnerable subpopulation will be inevitably23 infected at much higher rates, which could go up to 200% higher than under lockdown (Smith et al., 2022, 10–11), not only lacks evidence to back it up, but there is not even a mistaken assumption, such as asymptomatic transmission, to (supposedly) justify it! In my view, this is an important inconsistency which undercuts their rationale for it suggests that the model is based on very flimsy premises. Unsubstantiated claims are also made with respect to hospital capacity which implies that the circular reasoning of the Imperial modeling is a problem with this model too. The authors posit that hospitals are to be overflowed without strict mitigation measures and then calculate the impact of following a focused protection strategy on hospitalization rates and ICU admissions. The authors do not address that there is evidence suggesting that hospitalization rates are irrelevant to mitigation measures. Other postulates used in this model are also mishandled. The authors argue that even low risk groups could self-restrict their moves to avoid infection which could result in lower levels of population immunity and that this could leave room for future outbreaks (Smith et al., 2022, 11–12). Even though this is an accurate statement the way the authors treat it leads to them to highly questionable conclusions such as that this premise somehow speaks against focused protection strategies. As I have argued above, when low-risk groups voluntarily restrict their moves this is akin to a lockdown and it leads us to a situation whereby low-risk groups take fewer risks and high-risk groups are more exposed because the virus is still transmitted under (voluntary or mandatory) lockdown. Therefore, this is a failure of lockdowns that expose the elderly while delaying herd immunity, and thus prolonging the pandemic, not a failure of focused protection, as the authors contend. In a phrase, Smith et al. do not  It is of course important to mention that focused protection theorists do not claim (at least as far as I am concerned) that no vulnerable individual will be infected if such a policy is implemented, which is a claim that no reasonable individual can make, they simply argue that the vulnerable are better protected if the measures taken are age-specific instead of “one-size-fits-all.” Thus, what Smith et al. mention as “imperfect shielding,” which they take it to be a “critical weakness” of focused protection (Smith et al., 2022, 10), turns out to be a trivial observation, for no one ever claimed that targeted protection can be “perfect” in any meaningful way. The supposed “third critical weakness” of focused protection is also based on a trivial observation. The authors claim that even if herd immunity is achieved the vulnerable could still be infected because it cannot be formally excluded that the virus would be transmitted to them. This is correct, but it can happen with all endemic viruses; the OC43 outbreak in care homes in Canada is a typical example of this sort. It was an endemic coronavirus that entered care facilities and resulted in high death toll. Focused protection is a policy prescription that tries to minimize harm and to protect those at really high risk prior to endemicity; it does not make a case suggesting that the vulnerable will never catch again any endemic bug. It is therefore hard to see how the possibility that endemic viruses can infect individuals is a convincing argument against focused protection. 23

44

2  Viral Mitigation: Weak Theoretical Underpinnings

offer a solid account and their construction faces pretty much the same hurdles as nearly all covid-related models. A common pattern of pandemic modeling is that all of them rely on extremely dubious postulates that are usually not backed up by strong evidence and are not induced by sound scientific knowledge and then they construct a hypothetical framework and try to prove these very claims that lack empirical foundation. This strategy results in conclusions and outcomes that glaringly contradict common empirical and theoretical knowledge. One way to improve modeling is to make assumptions that are empirically grounded and considered as established knowledge, and then construct a hypothetical state of affairs and run algorithms to see how real-world data could perform under several scenarios. This could lead to more sound conclusions than the ones covid-modeling offered.

2.10 Displacement Effect, Lockdowns, and Focused Protection Imagine that societies were structured as they were functioning under lockdowns and that this was the norm. In this counterfactual, people live as if they are in a permanent state of lockdown: they are working from home, are studying from home, and they are leaving home only to visit supermarkets and to have a walk at the park nearby their houses and when they leave home they are always wearing masks. Based on the discussion throughout this chapter, we can now ask: do we have good reasons to believe that if people permanently lived like this they would not be exposed to pathogens that cause flu-like symptoms and, occasionally, serious disease? In other words, is there a social mechanism that can be put into place so that people will avoid coronaviruses, flu-like viruses, and a slew of bacteria and germs? If one answers “no,” then it can be argued that they have grasped the spinelessness of the reasoning behind the lockdown policies and that a convincing case against lockdowns has been made so far in this chapter. If one answers “yes,” then some clarifications should follow to try to convince them that they are in the wrong. The lockdown failures which we have discussed, as well as our theoretical knowledge on viruses, indicate that it is hardly the case that under such a lockdown-­ like social construction people would move in a microbe-free arena, and that the incidences of pneumonias would be significantly lower than they are under what is now regarded as common (no lockdown) conditions.24 If there is a lesson to be drawn by the lockdown failure is that pathogens circulate in the same places that (symptomatic) people move. So, if individuals move in supermarkets and in their houses only, transmission will take place in these settings. What is possible to be  Of course, if people did not travel at all, then the germs that circulated in one country could perhaps be absent from another which could lead to lower levels of cross-reactive immunity. But people in, say country A, would indeed be exposed to the pathogens that would be present in this place even if they followed lockdown protocols from the cradle to the grave. 24

2.10  Displacement Effect, Lockdowns, and Focused Protection

45

done though, at least for a period of time, is to try to isolate some groups of the population (the most vulnerable) and let others interact freely, as focused protection proponents argued. If all pathogens circulate with the same intensity, then the isolated group will be protected from all of them simultaneously while the others will likely be exposed to all of them in tandem. However, there is strong statistical support to the hypothesis that during the course of an epidemic, one pathogen prevails and the other endemic ones are restricted and circulate less, the so-called “displacement effect.” One reason for that could be that the pathogen that will circulate the most would be the one against which pre-existing immunity is not so prevalent. For example, influenza strains mutate much more than coronaviruses, adenoviruses, and other similar germs do, and so they are the ones that usually prevail for several months per year because pre-existing immunity against some of its epitopes seems to wane. While there are few data to support this claim and even if my assumption is correct, it may not be the only explanation (thus further research is needed to reveal the underlying mechanism that lies behind the displacement effect) it appears to be a matter of fact viruses25 interact in complex ways which results in common cold infections being less frequent when influenza strains dominate (Nickbakhsha et al., 2019). The displacement effect seems to have happened during the Covid-19 crisis as well. Sars-­ Cov-­2 dominated all other respiratory viruses for the best part of the pandemic as other respiratory viruses, especially the annual flu, did during the past years. The researchers used testing results and plotted the impact of several respiratory viruses over the years 2015–2020. A combined nose and throat swab was taken from all participants at recruitment and tested using a multiplex molecular respiratory virus panel for a range of targets which included influenza A, influenza B, Hcovs (HKU1, NL63, 229E, and OC43), human rhinovirus/enterovirus (hRV), adenovirus, human metapneumovirus (hMPV), parainfluenza virus (1, 2, 3, and 4), and respiratory syncytical virus (RSV). Sars-Cov-2 was added in 2020. Influenza A and B strongly dominated the years before 2020 (26% positive of the individuals tested), while hRV followed. Hcovs, hMPV, and parainfluenza were also detected. In 2020, it was Sars-Cov-2 that dominated, hRV was second, and other respiratory viruses have been in decline (Poole et al., 2020). The viral interference that leads to the displacement effect appears to explain why influenza viruses that commonly circulated in the population every winter were absent during 2020 which in turn debunks a lightweight argument of lockdown advocates according to which it was due to mitigation measures that influenza did not cause morbidity. If that was the case then they should develop an account which would explain why in the absence of lockdowns influenza dominated and other viruses were restricted prior to 2020. After that, they need to show us why Sars-­ Cov-­2 circulated under lockdown while other respiratory viruses did not. Since this

 Since Sars-Cov-2 is a virus, and since most available research on the displacement effect is related to viruses and not bacteria, the discussion here also focuses on viruses. 25

46

2  Viral Mitigation: Weak Theoretical Underpinnings

seems an unlikely prospect, we have solid reasons to put the absence of influenza strains in 2020 down to the displacement effect and not to lockdowns. The fact that it is one virus that circulates mostly each year could make us being relatively sure that the chances of most people to get infected are that this will happen with the prevailing virus and not with the ones that are in decline over that period of time. So, it is not the case that the risk of dying from Sars-Cov-2 is added to the already existing risks from other viruses, but it is the case that Sars-Cov-2 poses a risk almost on its own in the same way dominant influenza strains did until 2019. This reveals some worth-mentioning virus-related tradeoffs that are to be taken into account, especially when different policy prescriptions are on offer. Lockdowns were the prevailing policy when the pandemic started while focused protection appeared later in 2020 and challenged the conventional wisdom. There is also the “let it rip through” proposal which rejects even targeted measures for the elderly. By and large, the results of a “let it rip” approach would be better than a lockdown approach if the young are well informed and go on with their lives and the elderly take by themselves some precautions but could also be similar to a lockdown if all age-groups take voluntarily strong precautions. However, since it is mostly a principled stance that rejects governmental interventions of any sort and does not claim that it offers a “science-based” approach, I will not discuss it in detail. We will thus focus on how lockdowns and focused protection expose several age-groups to Sars-Cov-2 (which was the dominant virus), what kind of risk people face in each case, and whether the age-adjusted risk substantially differs when different viruses are dominant. As stated, under a lockdown schools and businesses are closed, most people work from home (apart from essential workers) and typically curfews are in place. People leave their houses for very specific reasons (a short walk or going to the grocery store). A focused protection policy suggests that the schools remain open, that businesses do so as well, and that healthy people below 65 years old go to work and do not work from home. Special protocols are to be followed for care homes and work-from-home is suggested for people over 65 years old and for people below 65  years old with very serious comorbidities (obese people are also included). People whose job is such that it cannot be done remotely, they could be involved in a furlough scheme. These measures26 would be in place until the virus reaches an endemic equilibrium which would make it difficult to infect the vulnerable, and when that happens, they would return to their normal lives. If pre-existing immunity is not high enough to prevent epidemic waves, herd immunity will be reached after two waves, and this is what happened in Sweden, where the first wave occurred in March to April 2020 and the second in December 2020 to January 2021. By January 2021 onwards, the pandemic has died out (FOHM) which suggest that robust

 I take for granted that the government has eligibility to intervene and I leave aside the ethical discussion on whether this should be the case. However, it should be stressed that the architects of focused protection strategy argue against medical mandates even if such mandates could be supported by the available data. 26

2.10  Displacement Effect, Lockdowns, and Focused Protection

47

population immunity had been built up. If pre-existing immunity is high, as it was in Asia, then even targeted measures are to a large extent useless. So, we know that it is Sars-Cov-2 that mainly circulates in 2020 and that influenza, which dominated every year before 2020, is absent. If schools are closed, then schoolchildren avoid a setting in which Sars-Cov-2 could be transmitted and in which various influenza strains circulated every year. As we have seen the IFR for schoolchildren is near to 0% and it is by now rather clear that influenza is more dangerous for schoolchildren than Sars-Cov-2 is (Yilmaz et al., 2021). So, schoolchildren would have been safer in school during 2020 when Sars-Cov-2 circulated than they were prior to that when it was the influenza strains that dominated. As for teachers, since they are typically below 65 years old, the risk of dying from Covid-19 which ranges from 0.023% for younger teachers (say in their thirties) 0.15% for teachers aged 50–59 and may go up to 0.2% for people 60–65 years old, which is in the same ballpark with the annual flu, since we saw that H1N1 has an IFR 0.1%, and it is generally accepted that this is the typical IFR for influenza. So, they do not face substantially higher risk of dying during 2020 than they did before that year. Furthermore, since, as we saw, schoolchildren do not transmit the virus that much as they did with influenza, it turns out that schools are a safe place for teachers too (Fenton et al., 2021). Thus, when a lockdown is in place children presumably avoid a risk that is lower than the one they faced prior to 2020 and teachers avoid a risk similar to that. When focused protection is in place, children are exposed to a negligible risk which is lower to the one they faced when they were exposed to influenza while teachers face pretty much similar risks both against influenza and against Sars-Cov-2 but given that schoolchildren transmit Sars-Cov-2 at lower rates than they transmitted influenza then the risk for teachers may as well be even lower to the one they faced in pre-pandemic years. So focused protection does not increase the risk of schoolchildren nor of teachers. When businesses are considered, we have similar results. Under lockdown, workers who are typically aged 18–65 supposedly avoid a risk of dying in the region of 0.010% for workers in their twenties to 0.15%-0.2% to those in their fifties to their early sixties, which is not significantly higher to the one they are usually exposed to. Under focused protection thus they face approximately the same risk with the one they were usually exposed to. Under focused protection thus they face approximately the same risk with the one they were usually exposed to. People over 65 face substantially higher risk of dying from Covid-19 than younger age-groups do, and although this group also faces higher risk of dying from influenza than younger people (Czaja et al., 2019), it is a matter of fact that for people above 65 Sars-Cov-2 is deadlier than the flu. So, during 2020, they could be exposed to a higher risk of dying than the one they were exposed when the flu circulated. Under lockdowns, they’ve been told that they avoid catching Sars-Cov-2. But we saw that this is not the case for they move in the same places that low-risk people move. Focused protection, on the contrary, requires that these people take the fewest possible risks for some period of time and the youngers go on with their lives so that the virus will circulate mostly in places were young people go and less to places where the vulnerable go. But if all move around the same

48

2  Viral Mitigation: Weak Theoretical Underpinnings

places such as houses, grocery stores, and the like, then the dominant virus will also be only in these places. The comparison discussed here suggests that in the presence of Sars-Cov-2 and in the absence of influenza, schools are safer for schoolchildren than they would be in the absence of Sars-Cov-2 and in the presence of influenza. We also see that the risk for most workers aged 18–65 is approximately the same whichever the dominant virus is. Thus, a no lockdown policy, whether it is constructed along the lines of the focused protection proposal or not, does not increase the daily risk of dying for sizeable segments of the population while it simultaneously increases the chances of survival for the remaining subpopulation that faces a higher risk in the absence of flu and in the presence of Sars-Cov-2. It thus appears that if people lived in a permanent state of lockdown, then when influenza was the prevailing virus, the elderly would be, again, even more exposed and the younger people would be more protected. The claim that a society can be germ-free is a fool’s errand. People cannot stop the spread nor eliminate the disease; they can only channel the epidemic wave towards those that face minuscule risk (if no interventions are in place) or towards those who face high risk (as it seems to happen under lockdowns). So, speaking of the Covid-19 crisis, if the purpose was to increase the chances of the vulnerable to avoid the virus, no lockdowns should have been in place, and if any measures were to be enforced, these should focused exclusively on the elderly. Obviously, I do not claim that protection for the elderly is guaranteed if the youngsters are exposed to the virus, but it appears that the chances of the vulnerable to get infected are reduced. After all, the higher death toll in countries that locked down indicates on its own that the vulnerable people were far from protected under lockdown.

References Abaluck, J., Kwong, L. H., Styczynski, A., Haque, A., et al. (2021). The impact of community masking on COVID-19: A cluster-randomized trial in Bangladesh. Science, 375(6577). https:// doi.org/10.1126/science.abi9069 Abu-Raddad, L.  J., Chemaitelly, H., Coyled, P., Malek, J.  A., Ahmed, A., Mohamoud, Y.  A., et  al. (2021). SARS-CoV-2 antibody-positivity protects against reinfection for at least seven months with 95% efficacy. EClinical Medicine, 35, 100861. https://doi.org/10.1016/j. eclinm.2021.100861 Ader, F., Bouscambert-Duchamp, M., Hites, M., Peiffer-Smadja, N., et  al. (2022). Remdesivir plus standard of care versus standard of care alone for the treatment of patients admitted to hospital with COVID-19 (DisCoVeRy): A phase 3, randomised, controlled, open-label trial. The Lancet, 22(2), 209–221. https://doi.org/10.1016/S1473-­3099(21)00485-­0 Aguas, R., Corder, R.  M., King, J.  G., & Gonçalves, G., et  al. (2020). Herd immunity thresholds for SARS-CoV-2 estimated from 2 unfolding epidemics. Pre-print at MedRix. https://doi. org/10.1101/2020.07.23.20160762 Altarawneh, H.  N., Chemaitelly, H., Hasan, M.  R., Ayoub, H.  H., Qassim, S., Coyle, P., et  al. (2022). Protection against the omicron variant from previous SARS-CoV-2 infection. The New England Journal of Medicine, 386, 1288–1290. https://doi.org/10.1056/NEJMc2200133

References

49

Atkinson, B., & Pedersen, E. (2020). Sars-Cov-2 shedding and infectivity. The Lancet, 395, 1339–1340. https://doi.org/10.1016/S0140-­6736(20)30868-­0 Australian Government, Department of Health. (2021). Coronavirus (COVID-19) at a glance. https://www.health.gov.au/resources/publications/coronavirus-­c ovid-­1 9-­a t-­a -­g lance-­6 -­ august-­2020. Accessed 6 Aug 2020. Axfors, C., & Ioannidis, J. (2022). Infection fatality rate of Covid-19  in community-dwelling elderly populations. European Journal of Epidemiology, 37, 235–249. https://doi.org/10.1007/ s10654-­022-­00853-­w Axfors, C., Schmitt, A.  M., Janiaud, P., Van’t Hooft, J., et  al. (2021). Mortality outcomes with hydroxychloroquine and chloroquine in COVID-19 from an international collaborative ­meta-­analysis of randomized trials. Nature Communications, 12, 2349. https://doi.org/10.1038/ s41467-­021-­22446-­z Baillie, V. E., Moore, D. P., Mathunjwa, A., Park, D. E., et al. (2021). Epidemiology and seasonality of endemic human coronaviruses in South African and Zambian children: A case-control pneumonia study. Viruses, 13(8), 1513. https://doi.org/10.3390/v13081513 Ballotpedia. (2021). School responses in South Dakota to the coronavirus (COVID-19) pandemic during the 2020–2021 school year. https://ballotpedia.org/School_responses_in_South_ Dakota_to_the_coronavirus_(COVID-­19)_pandemic_during_the_2020-­2021_school_year Barrett, S. (2004). Eradication versus control: The economics of global infectious disease policies. Bulletin of the World Health Organization, 82, 683–688. Barrett, S., & Hoel, M. (2007). Optimal disease eradication. Environmental and Development Economics, 12, 627–652. Barro, R. J. (2020). Non-pharmaceutical interventions and mortality in U.S. cities during the great influenza pandemic, 1918–1919 (Working paper no. 27049). National Bureau of Economic Research. https://doi.org/10.3386/w27049 Barro, R.  J., Ursua, J., & Weng, J. (2020). The coronavirus and the great influenza pandemic: Lessons from the “Spanish Flu” for the coronavirus’s potential effects on mortality and economic activity (Working paper no. 26866). National Bureau of Economic Research. https://doi. org/10.3386/w26866 Bendavid, E., Christopher, O., Bhattacharya, J., & Ioannidis, J. (2021a). Assessing mandatory stay-at-home and business closure effects on the spread of Covid-19. European Journal of Clinical Investigation, 51(6501). https://doi.org/10.1111/eci.13484 Bendavid, E., Mulaney, B., Sood, N., Shah, S., et al. (2021b). Covid-19 antibody seroprevalence in Santa Clara County, California. International Journal of Epidemiology, 50(2), 410–419. https://doi.org/10.1093/ije/dyab010 Bolourian, A., & Mojtahedi, Z. (2021). COVID-19 and flu pandemics follow a pattern: A possible cross-immunity in the pandemic origin and graver disease in farther regions. Archives of Medical Research, 52(2), 240–241. Booth, A., Reed, A.  B., Ponzo, S., Yasaee, A., et  al. (2021). Population risk factors for severe disease and mortality in Covid-19: A global systematic review and meta-analysis. PLoS One. https://doi.org/10.1371/journal.pone.0247461 Bower, H., Frisell, T., Di Giuzeppe, D., Delcoigne, B., et al. (2021). Impact of the COVID-19 pandemic on morbidity and mortality in patients with inflammatory joint diseases and in the general population: A nationwide Swedish cohort study. Epidemiological Science, 80, 1086–1093. Braun, J., Loyal, L., Frentsch, M., et al. (2020). SARS-CoV-2-reactive T cells in healthy donors and patients with COVID-19. Nature, 587, 270–274. https://doi.org/10.1038/s41586-­020-­2598-­9 Brauner, J.  M., Mindermann, S., Shamra, M., et  al. (2021). Inferring the Effectiveness of Government Interventions against Covid-19. Science, 371, 6531. https://doi.org/10.1126/science.abd9338 Bundgaard, H.,  Johan S.B., Daniel E.,  Tandeusz R. P. (2021). Effectiveness of adding a mask recommendation ot other public health measures to prevent Sars-Cov-2 infection in Danish mask-­wearers. Annals of Internal Medicine. https://doi.org/10.7326/M20-6817

50

2  Viral Mitigation: Weak Theoretical Underpinnings

Burgess, S., Ponsford, M. J., & Gill, D. (2020). Are we underestimating seroprevalence of SARS-­ CoV-­2? The British Medical Journal, 370, m3364. https://doi.org/10.1136/bmj.m3364 Caly, L., Druce, J. D., Catton, M. G., & Jans, D. A. (2020). The FDA-approved drug ivermectin inhibits the replication of SARS-CoV-2 in vitro. Antiviral Research, 178, 104787. https://doi. org/10.1016/j.antiviral.2020.104787 Campbell, D., & Stanley, J. (1963). Experimental and quasi-experimental design for research. Houghton Mifflin Company. Cao, S., Gan, Y., Wang, C., et  al. (2020). Post-lockdown SARS-CoV-2 nucleic acid screening in nearly ten million residents of Wuhan, China. Nature, 11, 5917. https://doi.org/10.1038/ s41467-­020-­19802-­w CDC. (2009). Serum cross-reactive antibody response to a novel influenza A (H1N1) virus after vaccination with seasonal influenza vaccine. Morbidity and Mortality Weekly Report, 58(19), 521–524. https://www.ncbi.nlm.nih.gov/pubmed/19478718.pmid:19478718. Accessed 22 May 2009. CDC. (2016a). What is smallpox? https://www.cdc.gov/smallpox/about/index.html. Accessed 7 June 2016. CDC. (2016b). Bioterrorism. https://www.cdc.gov/smallpox/bioterrorism/public/index.html. Accessed 19 Dec 2016. CDC. (2020a). Nonpharmaceutical measures for pandemic influenza in non-healthcare settings— Personal protective and environmental measures. https://wwwnc.cdc.gov/eid/article/26/5/19­0994_article. Accessed 16 Apr 2020. CDC. (2020b). Covid-view summary. https://www.cdc.gov/coronavirus/2019-­ncov/covid-­data/ covidview/past-­reports/05152020.html. Accessed 15 May 2020. CDC. (2020c). Covid-19 pandemic planning scenarios. https://www.who.int/mediacentre/news/ statements/2009/h1n1_pandemic_phase6_20090611/en/. Accessed 10 Sept 2020. CDC. (2021a). Outbreak of Sars-Cov-2 infections, including Covid-19 vaccine breakthrough infections, associated with large public gatherings, Barnstable County, Massachusetts, July 2021. https://www.cdc.gov/mmwr/volumes/70/wr/mm7031e2.htm?s_cid=mm7031e2_w. Accessed 30 July 2021. CDC. (2021b). Comparative effectiveness of Moderna, Pfizer-BioNTech, and Janssen (Johnson & Johnson) vaccines in preventing COVID-19 hospitalizations among adults without immunocompromising conditions—United States, March–August 2021. https://www.cdc.gov/mmwr/ volumes/70/wr/mm7038e1.htm?s_cid=mm7038e1_w. Accessed 24 Sept 2021. CDC. (2022a). Disease burden of flu. https://www.cdc.gov/flu/about/burden/index.html. Accessed 7 Jan 2022. CDC. (2022b). Life expectancy. https://www.cdc.gov/nchs/fastats/life-­expectancy.htm. Accessed 25 Mar 2022. CDC. (2022c). Weekly updates by select demographic and geographic characteristic. Provisional death counts for coronavirus disease 2019 (COVID-19). https://www.cdc.gov/nchs/nvss/vsrr/ covid_weekly/index.htm. Accessed 24 Aug 2022. Chaudhry, R., Dranitsaris, G., Mubashir, T., Bartoszko, J., & Riazi, S. (2020). A country level analysis measuring the impact of government actions, country preparedness and socioeconomic factors on COVID-19 mortality and related health outcomes. The Lancet, 25, 100464. https:// doi.org/10.1016/j.eclinm.2020.100464 Chemaitelly, H., Bertollini, R., & Abu-Raddad, L. J. (2021). National Study Group for COVID-19 Epidemiology. Efficacy of natural immunity against SARS-CoV-2 reinfection with the beta variant. The New England Journal of Medicine, 385, 2585–2586. Chernozhukov, V., Kasahar, H., & Schrimpf, P. (2021). Causal impact of masks, policies, behavior on early Covid-19 pandemic in the U.S. Journal of Econometrics, 220(1), 23–62. https://doi. org/10.1016/j.jeconom.2020.09.003 Chin, V., Ioannidis, J. P. A., Tanner, M. A., & Cripps, S. (2021). Effect estimates of COVID-19 non-pharmaceutical interventions are non-robust and highly model-dependent. Journal of Clinical Epidemiology, 136, 96–132. https://doi.org/10.1016/j.jclinepi.2021.03.014

References

51

Cohen, K. W., Linderman, S. L., Moodie, Z., Suthar, M. S., et al. (2021). Longitudinal analysis shows durable and broad immune memory after SARS-CoV-2 infection with persisting antibody responses and memory B and T cells. Cell Reports Medicine, 2(7), 100354. https://doi. org/10.1016/j.xcrm.2021.100354 Czaja, C.  A., Miller, L., Alden, N., Wald, H.  L., et  al. (2019). Age-related differences in hospitalization rates, clinical presentation, and outcomes among older adults hospitalized with influenza—U.S. Influenza Hospitalization Surveillance Network (FluSurv-NET). Open Forum Infectious Diseases, 6(7), 225. https://doi.org/10.1093/ofid/ofz225 d’Onofrio, A., & Manfredi, P. (2009). Information-related changes in contact patterns may trigger oscillations in the endemic prevalence of infectious diseases. Journal of Theoretical Biology, 256(3), 473–478. Damon, I. K., Damaso, C., & Fadden, G. M. (2014). Are we there yet? The smallpox research agenda using variola virus. Plos Pathogens, 10(5), e1004108. https://doi.org/10.1371/journal. ppat.1004108 Doshi, P. (2020a). Covid-19: Do many people have pre-existing immunity? The British Journal of Medicine, 370, m3563. https://doi.org/10.1136/bmj.m3563 Doshi, P. (2020b). Will Covid-19 vaccines save lives? Current trials aren’t designed to tell us. The British Medical Journal, 371, m4037. https://doi.org/10.1136/bmj.m4037 Doshi, P. (2021). Pfizer and Moderna’s “95% effective” vaccines—We need more details and the raw data. The British Medical Journal. https://blogs.bmj.com/bmj/2021/01/04/peter-­doshi-­ pfizer-­and-­modernas-­95-­effective-­vaccines-­we-­need-­more-­details-­and-­the-­raw-­data/ Dougan, M., Nirula, A., Azizad, M., Mocherla, B., et al. (2021). Bamlanivimab plus etesevimab in mild or moderate Covid-19. The New England Journal of Medicine, 385, 1382–1392. https:// doi.org/10.1056/NEJMoa2102685 ECDC. (2021). Using face masks in the community: First update. https://www.ecdc.europa.eu/ sites/default/files/documents/covid-­19-­face-­masks-­community-­first-­update.pdf. Accessed 15 Feb 2021. Edridge, A.  E. D., Kaczorowska, J., Hoste, A.  C. R., Bakker, M., et  al. (2020). Seasonal coronavirus protective immunity is short-lasting. Nature Medicine, 26, 1691–1693. https://doi. org/10.1038/s41591-­020-­1083-­1 FDA. (2020). FDA approves first treatment for Covid-19. https://www.fda.gov/news-­events/press-­ announcements/fda-­approves-­first-­treatment-­covid-­19. Accessed 22 Oct 2020. FDA Briefing Document. (2020). Pfizer-BioNTech COVID-19 vaccine. https://www.fda.gov/ media/144245/download#page=42. Accessed 10 Dec 2020. Fenton, L., Gribben, C., Caldwell, D., Colville, S., et al. (2021). Risk of hospital admission with Covid-19 among teachers compared with healthcare workers and other adults of working age in Scotland, March 2020 to July 2021: Population based case-control study. The British Medical Journal, 374, n2060. https://doi.org/10.1136/bmj.n2060 Ferguson, N. M., Cummings, D. A. T., Cauchemez, S., et al. (2005). Strategies for containing an emerging influenza pandemic in Southeast Asia. Nature, 437, 209–214. https://doi.org/10.1038/ nature04017 Ferguson, N. M., Cummings, D. A. T., Fraser, C., et al. (2006). Strategies for containing an influenza pandemic. Nature, 442, 448–452. https://doi.org/10.1038/nature04795 Ferguson, N.  M., Daniel, L., Gemma, N.-G., Natusko, I., Kylie, A., Marc, B., Sangeeta, B., et  al. (2020). Report 9: Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand. Imperial College. https://www.imperial. ac.uk/media/imperial-­college/medicine/sph/ide/gidafellowships/Imperial-College-COVID19NPI-­modelling-­16-­03-­2020 Fidgerald, D. (2009). Human swine influenza a [H1N1]: Practical advice for clinicians early in the pandemic. Pediatric Respiratory Reviews, 10(3), 154–158. https://doi.org/10.1016/j. prrv.2009.06.005 Flaxman, S., Mishra, S., Gandy, A., et al. (2020). Estimating the effects of non-­pharmaceutical interventions on Covid-19  in Europe. Nature, 584, 257–261. https://doi.org/10.1038/ s41586-­020-­2405-­7

52

2  Viral Mitigation: Weak Theoretical Underpinnings

FOHM.  Covid-19. https://experience.arcgis.com/experience/09f821667ce64bf7be6f9f87457ed 9aa/page/page_0/ For a Healthy Belgium. (2021). Life expectancy. https://www.healthybelgium.be/en/health-­status/ life-­expectancy-­and-­quality-­of-­life/life-­expectancy. Accessed 19 Sept 2021. Gao, M., Jang, L., Chen, X., Deng, Y., et al. (2020). A study on infectivity of asymptomatic Sars-­ Cov-­2 carriers. Respiratory Medicine. https://doi.org/10.1016/j.rmed.2020.106026 Gardner, J., Willem, L., & Van Der Wijngaart, W., et  al. (2020). Intervention strategies against Covid-19 and their estimated impact on Swedish healthcare capacity. Preprint at MedRix. https://doi.org/10.1101/2020.04.11.20062133 Gilca, R., Carazo, S., Amini, R., Charest, H., & De Serres, G. (2020). Relative severity of common human coronaviruses and influenza in patients hospitalized with acute respiratory infection: Results from 8-year hospital-based surveillance in Quebec, Canada. The Journal of Infectious Diseases, 223(6), 1078–1087. Great Barrington Declaration. (2020). https://gbdeclaration.org/. Accessed 4 Oct 2020. Grifoni, A., Weiskopf, D., Ramirez, S.  I., & Mateus, J. (2020). Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals. Cell, 181(7), 1489–1501. https://doi.org/10.1016/j.cell.2020.05.015 Gudbjartsson, D. F., Norddahl, G. L., Melsted, P., Gunnarsdottir, K., Holm, H., Eythorsson, E., Arnthorsson, A.  O., et  al. (2020). Humoral immune response to SARS-CoV-2  in Iceland. New England Journal of Medicine, 383, 1724–1734. https://www.nejm.org/doi/full/10.1056/ NEJMoa2026116 Guerra, D., & Guerra, D. (2021). Mask mandate and use efficacy for Covid-19 containment in US states. International Research Journal of Public Health, 5, 55. https://doi.org/10.28933/ irjph-­2021-­08-­1005 Hall, R.  G. (2011). Political and social determinants of disease eradication. In S.  L. Cochi & W. R. Dowdle (Eds.), Disease eradication in the 21st century (pp. 47–61). MIT Press. Harrison, E.  G., & Kuo, L. (2020). China’s coronavirus lockdown-strategy: Brutal but effective. The Guardian. https://www.theguardian.com/world/2020/mar/19/chinas-­coronavirus-­ lockdown-­strategy-­brutal-­but-­effective. Accessed 19 Mar 2020. Haug, N., Geyrhofer, L., Londei, A., Dervic, E., Desvars-Larrive, A., Loreto, V., Pinior, B., et al. (2020). Ranking the effectiveness of worldwide COVID-19 government interventions. Nature Human Behavior, 4, 1303–1312. He, X., Lau, E.H.Y., Wu, P. et.al. (2020). Temporal dynamics in viral shedding and transmissibility of COVID-19. Nature Medicine. 26, 672–675. Henderson, D. A., Inglesby, T. V., Bartlett, J. G., et al. (1999). Smallpox as a biological weapon: Medical and public health management. Journal of the American Medical Association, 281(22), 2127–2137. Henderson, D. A., Courtney, B., Inglesby, T. V., et al. (2009). Public health and medical responses to the 1957–58 influenza pandemic. Biosecurity and Bioterrorism: Biodefense Strategy, Practice and Science, 7(3), 265–273. Herby, J., Jonung, L., & Hanke, S. H. (2022, May). A literature review and meta-analysis of the effects of lockdowns on Covid-19 mortality-II. Studies in Applied Economics, 210. Hinman, A.  R., & Hopkins, D.  R. (1998). Lessons from previous eradication programs. In W. R. Dowdle & D. R. Hopkins (Eds.), The eradication of infectious diseases (Dahlem workshop report, J. Lupp, Series Ed.) (pp. 19–31). John Wiley & Sons. Hou, Y. J., Okuda, K., Edwards, C.E., Martinez D.R. et.al. (2020). Sars-Cov-2 reverse genetics reveals a variable infection gradient in the respiratory tract. Cell (2): 429–446. e14. Howard, J., Huang, A., Li, Z., et al. (2020). Face masks against COVID-19: An evidence review. Preprints. https://doi.org/10.20944/preprints202004.0203.v1 Howard, J., Huang, A., Li, Z., et al. (2021). Face masks against COVID-19: An evidence review. PNAS, 118(4), e2014564118. https://doi.org/10.1073/pnas.2014564118 Inglesby, T. V., Nuzzo, J. B., O’Toole, T., & Henderson, D. A. (2006). Disease mitigation measures in the control of pandemic influenza. Biosecurity and Bioterrorism: Biodefense Strategy, Practice, and Science, 4(4), 366–375. https://doi.org/10.1089/bsp.2006.4.366

References

53

Ioannidis, J. (2020). Infection fatality rate of COVID-19 inferred from seroprevalence data. Bulletin of the World Health Organization. https://www.who.int/bulletin/online_first/ BLT.20.265892.pdf?fbclid=IwAR2OP5ZPKsYTGIZ2D1TWR3qp11yH87JXUn8Mogx XyC6-­iKDZnimP6jEzsr0 Ioannidis, J. (2021). Reconciling estimates of global spread and infection fatality rates of COVID-19: An overview of systematic evaluations. European Journal of Clinical Investigation, 51(5), e13554. https://doi.org/10.1111/eci.13554 Ioannidis, J., Crips, S., & Tanner, M. (2020). Forecasting for Covid-19 has failed. International Journal of Forecasting, 38(2), 423–438. https://doi.org/10.1016/j.ijforecast.2020.08.004 Jans, D.  A., & Wagstaff, K.  M. (2020). The broad spectrum host-directed agent ivermectin as an antiviral for SARS-CoV-2? Biomedical and Biochemical Research Communications, 538, 163–172. https://doi.org/10.1016/j.bbrc.2020.10.042 Jefferson, T., Heneghan, C., Spencer, E., & Brassey, J. (2020a). Are you infectious if you have a positive PCR test for Covid-19? The Center for Evidence Based Medicine. https://www.cebm. net/covid-­19/infectious-­positive-­pcr-­test-­result-­covid-­19/. Accessed 5 Aug 2020. Jefferson, T., Spencer, E.  A., Brassey, J., & Heneghan, C. (2020b). Viral cultures for Covid-19 infectious potential assessment  – A systematic review. Clinical Infectious Diseases, 73(11), e3884–e3899. https://doi.org/10.1093/cid/ciaa1764 Jefferson, T., Del Mar, C.  B., Dooley, L., et  al. (2020c). Physical interventions to interrupt or reduce the spread of respiratory viruses (review). Cochrane Database of Systematic Reviews. https://doi.org/10.1002/14651858.CD006207.pub5. Accessed 20 Nov 2020. Jung, J. H., Rha, M.-S., Sa, M., & Choi, H. K. (2021). SARS-CoV-2-specific T-cell memory is sustained in COVID-19 convalescent patients for 10 months with successful development of stem cell-like Memory T-Cells. Nature Communications, 12, 4043. https://doi.org/10.1038/ s41467-­021-­24377-­1 Kamerlin, S., & Kasson, P. (2020). Managing coronavirus disease 2019 spread with voluntary public health measures: Sweden as a case study for pandemic control. Clinical Infectious Diseases, 71(12), 3174–3181. https://doi.org/10.1093/cid/ciaa864 Karadimas, P. (2022a). Thought experiments and the pragmatic nature of explanation. Foundations of Science. https://doi.org/10.1007/s10699-­022-­09844-­2 Karadimas, P. (2022b). Covid-19, public policy and public choice theory. The Independent Review: A Journal of Political Economy, 27, (2), 273–302. Kuhbandner, C., Homburg, S., Walach, H., & Hockertz, S. (2022). Was Germany’s lockdown in spring 2020 necessary? How bad data quality can turn a simulation into a delusion that shapes the future. Futures, 135, 102879. https://doi.org/10.1016/j.futures.2021.102879 Kupferschmidt, K. (2020). Study claiming new coronavirus can be transmitted by people without symptoms was flawed. Science. https://www.science.org/content/article/ paper-­non-­symptomatic-­patient-­transmitting-­coronavirus-­wrong Larochelambert, Q., Andy, M., Antero, J., Le Bourg, E., & Toussaint, J. F. (2020). Covid-19 mortality: A matter of vulnerability among nations facing limited margins of adaptation. Frontiers of Public Health, 39, 6043. https://doi.org/10.3389/fpubh.2020.604339 Lavine, J.  S., Bjornstad, O.  N., & Antia, R. (2021). Immunological characteristics govern the transition of Covid-19 to endemicity. Science, 371(65301), 741–7545. https://doi.org/10.1126/ science.abe6522 Le Bert, N., Clapham, H. E., Tan, A. T., Chia, W. N., et al. (2021). Highly functional virus-specific cellular immune response in asymptomatic SARS-CoV-2 infection. Journal of Experimental Medicine, 218(5), e20202617. https://doi.org/10.1084/jem.20202617 Liu, I. T., Prasad, V., & Darrow, J. D. (2021). Evidence for community cloth face masking to limit the spread of SARS-CoV-2: A critical review (Working paper no. 64). CATO Institute. Luo, L., Liu, D., Liao, X., Wu, X., Jing, Q., Zheng, J., Liu, F., et al. (2020). Contact settings and risk for transmission in 3410 close contacts of patients with COVID-19 in Guangzhou, China. Annals of Internal Medicine, 173(11), 879–887. https://doi.org/10.7326/M20-­2671 Macrotrends. (2021). Australia life expectancy 1950–2021. https://www.macrotrends.net/ countries/AUS/australia/life-­expectancy

54

2  Viral Mitigation: Weak Theoretical Underpinnings

Mak, T., & Saunders, M. E. (2006). Innate immunity. In The immune response: Basic and clinical principles (pp. 69–92). Academic Press. Markel, H., Lipmann, H., & Navarro, A. (2007). Nonpharmaceutical interventions implemented by US cities during the 1918–1919 influenza pandemic. JAMA Network., 298(6), 644–654. Mateus, J., Alba, G., Tarke, A., et al. (2020). Selective and cross-reactive Sars-Cov-2 T cell epitopes in unexposed humans. Science, 370(6512), 89–94. https://doi.org/10.1126/science.abd3871 Matta, J., Wiernik, E., Robinseau, O., et al. (2021). Association of self-reported COVID-19 infection and SARS-CoV-2 serology test results with persistent physical symptoms among French adults during the COVID-19 pandemic. JAMA Internal Medicine, 182(1), 19–25. https://doi. org/10.1001/jamainternmed.2021.6454 McIntyre, R., Seale, H., Dang, T. S., et al. (2015). A cluster randomized trial of cloth masks compared with medical masks in healthcare workers. British Medical Journal, 5(4), e006577. https://doi.org/10.1136/bmjopen-­2014-­006577 Mina, M., Peto, T., Finana, M. G., Semple, M., & Buchan, I. E. (2021). Clarifying the evidence on Sars-Cov-2 antigen rapid tests in public health responses to Covid-19. The Lancet, 397(10282), 1425–1427. https://doi.org/10.1016/S0140-­6736(21)00425-­6 Moderna. (2020). Moderna announces primary efficacy analysis in phase 3 COVE study for its COVID-19 vaccine candidate and filing today with U.S. FDA for emergency use authorization. https://investors.modernatx.com/news-­releases/news-­release-­details/moderna-­announces-­ primary-­efficacy-­analysis-­phase-­3-­cove-­study. Accessed 30 Nov 2020. Mokhtari, M., Mohraz, M., Gouya, M. M., et al. (2021). Clinical outcomes of patients with mild Covid-19 following treatment with hydroxychloroquine in an outpatient setting. International Immunopharmacology, 96, 107636. https://doi.org/10.1016/j.intimp.2021.107636 Mondeli, M., Colaneri, M., Seminari, E.  M., Baldanti, F., & Bruno, R. (2020). Low risk of SARS-CoV-2 transmission by fomites in real-life conditions. Lancet, 21(5), e112. https://doi. org/10.1016/S1473-­3099(20)30678-­2 Mouton, C., Bazadua, O. V., Piercedavid, B., & Espino, V. (2001). Common infections in older adults. American Family Physician, 63(2), 257–269. Munro, A. P. S., & Faust, S. N. (2020). Children are not Covid-19 super spreaders: Time to go Back to school. British Journal of Medicine, 105, 618–619. https://doi.org/10.1136/archdisch ild-­2020-­319866 Murchu, E. O., Byrne, P., Carty, P. G., De Gascun, C., et al. (2021). Quantifying the risk of Sars-­ Cov-­2 reinfection over time. Reviews in Medical Biology, 32, 4. https://doi.org/10.1002/ rmv.2260 Ndam, J. (2020). Modelling the impacts of lockdown and isolation on the eradication of COVID-19. BioMath, 9(2). https://doi.org/10.11145/j.biomath.2020.09.107 Ng, K. W., Faulkner, N., Cornish, G. H., Rosa, A., et al. (2020). Preexisting and de novo humoral immunity to SARS-CoV-2 in humans. Science, 370, 1339–1343. Nickbakhsha, S., Mair, C., Mathews, L., et al. (2019). Virus–virus interactions impact the population dynamics of influenza and the common cold. PNAS, 116(52), 27142–27150. https://www. pnas.org/cgi/doi/10.1073/pnas.1911083116 Onakpoya, I. J., Heneghan, C. J., Spencer, E. A., Brassey, J., et al. (2021). SARS-CoV-2 and the role of fomite transmission: A systematic review. F1000Research. https://doi.org/10.12688/ f1000research.51590.2 ONS. (2021a). Average age of those who had died with Covid-19. https://www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/averageageofthosewhohaddiedwithcovid19. Accessed 11 Jan 2021. ONS. (2021b). Technical article: Updated estimates of the prevalence of post-acute symptoms among people with coronavirus (COVID-19) in the UK: 26 April 2020 to 1 August 2021. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/articles/technicalarticleupdatedestimatesoftheprevalenceofpostacutesymptomsamon gpeoplewithcoronaviruscovid19intheuk/26april2020to1august2021?fbclid=IwAR09zLgjC2_ YsHYqVsc4YOKgA4stqHDYnTEl75oRneJYQZrCx7HnOZNs_mM. Accessed 16 Sept 2021.

References

55

OWD. (2021a). Coronavirus pandemic data explorer. Sweden and South Korea. https://ourworldindata.org/explorers/coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..lat est&facet=none&pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirm ed+deaths&Interval=7-­day+rolling+average&Relative+to+Population=true&Align+outbreaks =false&country=KOR~SWE. Accessed 10 Aug 2021. OWD. (2021b). Coronavirus pandemic data explorer. Japan, US and EU. https://ourworldindata. org/explorers/coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..latest&fa cet=none&pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirmed+de aths&Interval=7-­day+rolling+average&Relative+to+Population=true&Align+outbreaks=false &country=JPN~European+Union~USA. Accessed 10 Aug 2021. OWD. (2021c). Coronavirus pandemic data explorer. Peru. https://ourworldindata.org/explorers/ coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..latest&facet=none& pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirmed+cases&Inter val=7-­day+rolling+average&Relative+to+Population=true&Align+outbreaks=false&country =~PER. Accessed 21 Sept 2021. OWD. (2021d). Coronavirus pandemic data explorer. New Zealand. https://ourworldindata.org/ explorers/coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..latest&facet= none&pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirmed+cases& Interval=7-­day+rolling+average&Relative+to+Population=true&Align+outbreaks=false&cou ntry=~NZL. Accessed 3 Oct 2021. OWD. (2021e). Coronavirus pandemic data explorer. Australia. https://ourworldindata.org/ explorers/coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..latest&facet= none&pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirmed+cases& Interval=7-­day+rolling+average&Relative+to+Population=true&Align+outbreaks=false&cou ntry=~AUS. Accessed 3 Oct 2021. OWD. (2021f). Coronavirus pandemic data explorer. Greece. https://ourworldindata.org/explorers/coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..latest&facet=none &pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirmed+deaths&Inte rval=7-­day+rolling+average&Relative+to+Population=true&Align+outbreaks=false&country =~GRC. Accessed 26 Dec 2021. OWD. (2022a). Coronavirus pandemic data explorer. Peru, UK, Italy, Belgium, France, Sweden. https://ourworldindata.org/explorers/coronavirus-­data-­explorer?zoomToSelection=true&t ime=2020-­03-­01..latest&pickerSort=desc&pickerMetric=total_deaths_per_million&Metric= Confirmed+deaths&Interval=7-­day+rolling+average&Relative+to+Population=true&Align+o utbreaks=false&country=SWE~PER~BEL~GBR~ITA~FRA. Accessed 25 Aug 2022. OWD. (2022b). Coronavirus pandemic data explorer. EU, Europe, Sweden. https://ourworldindata.org/explorers/coronavirus-­data-­explorer?zoomToSelection=true&time=2020-­03-­01..lates t&facet=none&pickerSort=desc&pickerMetric=total_deaths_per_million&Metric=Confirmed +deaths&Interval=7-­day+rolling+average&Relative+to+Population=true&Color+by+test+po sitivity=false&country=SWE~European+Union~Europe. Accessed 26 Aug 2022. Paixao, J., Gualanque, M., Gaston, C., Carralero, R., et  al. (2022). Early evidence of circulating SARS-CoV-2 in unvaccinated and vaccinated measles patients, September 2019–February 2020. Dovepress, 15, 533–544. https://doi.org/10.2147/IDR.S344437 Pape, K., Dileepan, T., Kabage, A. J., & Schacker, T. W. (2021). High-affinity memory B-cells induced by Sars-Cov-2 infection produce more plasmablasts and atypical memory B-cells than those primed by mRNA vaccines. Cell Reports, 37(2), 10923. https://doi.org/10.1016/j. celrep.2021.109823 Patrick, D.  M., Petric, M., Skowronski, D., et  al. (2006). An outbreak of human coronavirus OC43 infection and serological cross-reactivity with SARS coronavirus. Canadian Journal of Infectious Diseases and Medical Microbiology, 17(6), 330–336. Patrozou, E., & Mermel, L.  A. (2009). Does influenza transmission occur from asymptomatic infection or prior to symptom onset? Public Health Reports, 124, 193–195. Pfizer Biontech. (2020). Pfizer-Biontech Covid-19 vaccine (BNT162, PF-07302048). https://www. fda.gov/media/144246/download. Accessed 30 Nov 2020.

56

2  Viral Mitigation: Weak Theoretical Underpinnings

Polack, F.  P., Thomas, S.  J., Kitchin, N., Absalon, J., et  al. (2020). Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. The New England Journal of Medicine, 383, 2603–2615. https://doi.org/10.1056/NEJMoa2034577 Poole, S., Brendish, N. J., & Clark, T. W. (2020). SARS-CoV-2 has displaced other seasonal respiratory viruses: Results from a prospective cohort study. Journal of Infection, 81(6), 966–971. https://doi.org/10.1016/j.jinf.2020.11.010 Popp, M., Stegemann, M., Metzendorf, M.  I., Gould, S., et  al. (2021). Ivermectin for preventing and treating COVID-19. The Cochrane Database of Systematic Reviews. https://doi. org/10.1002/14651858.CD015017.pub2. Accessed 28 July 2021. Pratha, S., Fitzpatrick, M., Zimmer, C., et al. (2021). Asymptomatic Sars-Cov-2 infection: A systematic review and meta-analysis. PNAS, 118(34), e2109229118n. https://doi.org/10.1073/ pnas.2109229118 Prodromos, C., & Rumschlag, T. (2020). Hydroxychloroquine is effective, and consistently so when provided early, for COVID-19: A systematic review. New Microbes and New Infections, 38, 100776. https://doi.org/10.1016/j.nmni.2020.100776 Radonovich, L.  J., Simberkoff, M.  S., Bessesen, M.  T., et  al. (2019). N95 respirators vs medical masks for preventing influenza among health care personnel a randomized clinical trial. Journal of the American Medical Association, 322(9), 824–833. https://doi.org/10.1001/ jama.2019.11645 Rotevatn, T. A., Elstrom, P., Greve-Isdahl, M., Suren, P., et al. (2022). School closure versus targeted control measures for SARS-CoV-2 infection. Pediatrics, 149(5), e2021055071. https:// doi.org/10.1542/peds.2021-­055071 Rothe, C., Schunk, M., Sothmann, P., & Bretzel, G. (2020). Transmission of 2019-nCoV infection from an asymptomatic contact in Germany. The New England Journal of Medicine, 382, 970–971. https://doi.org/10.1056/NEJMc2001468 SAcoronavirus. (2021). Update on Covid-19. https://sacoronavirus.co.za/2020/06/24/update-­on-­ covid-­19-­24th-­june-­2020/. Accessed 24 Oct 2020. Sekine, T., Perez-Potti, A., Rivera-Ballesteros, O., Strålin, K., Gorin, J.-B., Olsson, A., Llewellyn-­ Lacey, S., et al. (2020). Robust T-cell immunity in convalescent individuals with asymptomatic or mild COVID-19. Cell, 183, 158–168. Sierra, B., Bossuyt, N., Braeye, T., Leroy, M., et  al. (2020). All-cause mortality supports the COVID-19 mortality in Belgium and comparison with major fatal events of the last century. Archives of Public Health, 78, 117. https://doi.org/10.1186/s13690-­020-­00496-­x Skountzou, I., Koutsonanos, D. G., Kim, J. H., Powers, R., Satyabhama, L., Masseoud, F., et al. (2010). Immunity to Pre-1950 H1N1 influenza viruses confers cross-protection against the pandemic swine-origin 2009 a (H1N1) influenza virus. The Journal of Immunology, 185(3), 1642–1649. https://doi.org/10.4049/jimmunol.1000091 Smith, C.  A., Yates, C.  A., & Ashby, B. (2022). Critical weaknesses in shielding strategies for Covid-19. PLOS Global Public Health. https://doi.org/10.1371/journal.pgph.0000298 Sridhar, S., Begom, S., Bermingham, A., Hoschler, K., Adamson, W., Carman, W., et al. (2013). Cellular immune correlates of protection against symptomatic pandemic influenza. Nature Medicine, 19, 1305–1312. https://doi.org/10.1038/nm.3350 Stang, A., Roberts, J., Schonert, B., Jockel, K. H., et al. (2021). The performance of the Sars-Cov-2 RT-PCR test as a tool for detecting Sars-Cov-2 infection in the population. Journal of Infection, 83(2), 237–279. https://doi.org/10.1016/j.jinf.2021.05.022 Statista. (2021). Death rates from coronavirus (COVID-19) in the United States as of December 16, 2021, by state. https://www.statista.com/statistics/1109011/coronavirus-­covid19-­death-­ rates-­us-­by-­state/. Accessed 19 Aug 2021. Steinman, J. B., Lum, F. M., Ho, P. P.-K., Kaminskid, N., & Steinman, L. (2020). Reduced development of Covid-19  in children reveals molecular checkpoints gating pathogenesis illuminating potential therapeutics. Perspective, 117(40), 24620–24626. https://doi.org/10.1073/ pnas.2012358117 Strausbaugh, L. J., Sukumar, S. R., Joseph, C. L., & High, K. P. (2003). Infectious disease outbreaks in nursing homes: An unappreciated hazard for frail elderly persons. Clinical Infectious Diseases, 36(7), 870–876.

References

57

Subramanian, S. V., & Kumar, A. (2021). Increases in Covid-19 unrelated to levels of vaccination across 68 countries and 1,947 counties in the United States. European Journal of Epidemiology, 36, 1237–1240. https://doi.org/10.1007/s10654-­021-­00808-­7 Sudre, C. H., Murray, B., Varsavsky, T., Graham, M. S., et al. (2020). Attributes and predictors of long-covid. Nature Medicine, 27, 626–631. https://doi.org/10.1038/s41591-­021-­01292-­y Tan, C., Lam, S. D., Richard, D., & Owen, C. J. (2022). Transmission of Sars-Cov-2 from humans to animals and potential host adaptation. Nature Communications, 13, 2988. https://doi. org/10.1038/s41467-­022-­30698-­6 Tarke, A., Sidney, J., Methot, N., Crotty, S., Griffonni, A., Sette, A., et al. (2021). Impact of SARS-­ CoV-­2 variants on the total CD4+ and CD8+ T cell reactivity in infected or vaccinated individuals. Cell Reports, 2(7), 100355. https://doi.org/10.1016/j.xcrm.2021.100355 Tartof, S.  Y., Jeff M.S., Heidi, F.  Vennis, H. et.al. (2021). Effectiveness of mRNA BNT162b2 COVID-19 vaccine up to 6 months in a large integrated health system in the USA: a retrospective cohort study. The Lancet. 1407-1416. https://doi.org/10.1016/S01140-6736(21)02183-8 Taub, D., Eshler, W., Janowski, M., Ferrucci, L., et  al. (2008). Immunity from smallpox vaccine persists for decades: A longitudinal study. The American Journal of Medicine, 121(12), 1058–1064. https://doi.org/10.1016/j.amjmed.2008.08.019 Tichopád, A., Pecen, L., & Sedlák, V. (2021). Could the new coronavirus have infected humans prior November 2019? PLoS One, 16(8), e0248255. https://doi.org/10.1371/journal.pone. 0248255 Tricado, J. L. S., Gomez-Perosanz, M., & Reche, P. A. (2017). Fundamentals and methods for Tand B-cell epitope prediction. Journal of Immunology Research. Article ID: 2680160. https:// doi.org/10.1155/2017/2680160 Turner, J. S., Kim, W., Kalaidina, E., Goss, C. W., et al. (2021). SARS-CoV-2 infection induces long-lived bone marrow plasma cells in humans. Nature, 595, 421–425. https://doi.org/10.1038/ s41586-­021-­03647-­4 UK Gov. Department of Health. (2011). UK influenza pandemic preparedness strategy 2011. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/ file/213717/dh_131040.pdf. Accessed 10 Nov 2011. UK Government and Parliament. (2021). Adopt a zero covid strategy – Full lockdown and close borders. https://petition.parliament.uk/petitions/571965 Unite Against Covid. (2020). History of the Covid-19 alert system. https://covid19.govt.nz/ alert-­levels-­and-­updates/history-­of-­the-­covid-­19-­alert-­system/ Unwin, H. J. T., Mishra, S., Bradley, V. C., et al. (2020). State-level tracking of Covid-19 in the United States. Nature Communications, 11, 6189. https://doi.org/10.1038/s41467-­020-­19652-­6 Van Fraassen, B. C. (1980). Scientific image. Oxford University Press. Varghese, S. (2020). Scotland is aiming to eliminate coronavirus, why not England? Wired. https:// www.wired.co.uk/article/zero-­covid-­coronavirus-­elimination. Accessed 8 Mar 2020 Vitale, J., Mumoli, N., Clerici, P., & De Paschale, M. (2021). Assessment of SARS-CoV-2 reinfection 1 year after primary infection in a population in Lombardy, Italy. JAMA Internal Medicine, 181(10), 1407–1408. https://doi.org/10.1001/jamainternmed.2021.2959 WHO. (2019). Global Influenza Program. Non-pharmaceutical public health measures for mitigating the risk and impact of epidemic and pandemic influenza. http://apps.who.int/iris/bitstream/handle/10665/329438/9789241516839-­eng.pdf WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19) 16–24 February. (2020a). Geneva: World Health Organization. https://www.who.int/publications/i/item/report-­of-­the-­ who-­china-­joint-­mission-­on-­coronavirus-­disease-­2019-­(covid-­19). Accessed 10 May 2020. WHO Director-General’s Opening Remarks at the Media Briefing on COVID-19 – 3 March. (2020b). Geneva: World Health Organization. https://www.who.int/director-­general/speeches/detail/ who-­director-­general-­s-­opening-­remarks-­at-­the-­media-­briefing-­on-­covid-­19%2D%2D-­3-­ march-­2020. Accessed 10 May 2020. WHO. (2020c). WHO recommends against the use of remdesivir in COVID-19 patients, https:// www.who.int/news-­r oom/feature-­s tories/detail/who-­r ecommends-­a gainst-­t he-­u se-­o f-­ remdesivir-­in-­covid-­19-­patients. Accessed 20 Nov 2020.

58

2  Viral Mitigation: Weak Theoretical Underpinnings

WHO. Smallpox. https://www.who.int/health-­topics/smallpox#tab=tab_1 WHO Director-General’s Statement to the Press. (2009). World now at the start of 2009 influenza pandemic. https://www.who.int/mediacentre/news/statements/2009/h1n1_pandemic_ phase6_20090611/en/. Accessed 11 June 2009. Wikipedia. (2022). Covid-19 pandemic in Peru. https://en.wikipedia.org/wiki/COVID-­19_pandemic_in_Peru#Response. Accessed 17 July 2022. Wilson, N., Mansoor, O. D., Boyd, M. J., Kvalsig, M., et al. (2021). We should not dismiss the possibility of eradicating Covid-19: Comparisons with smallpox and polio. BMJ Global Health, 6, e006810. https://doi.org/10.1136/bmjgh-­2021-­006810 Wood, S. (2020). Inferring UK COVID-19 fatal infection trajectories from daily mortality data: Were infections already in decline before the UK lockdowns? Biometrics, 30. https://doi. org/10.1111/biom.13462 World Atlas. (2019). Life expectancy in Asia. https://www.worldatlas.com/articles/asian-­countries-­ where-­people-­live-­the-­longest.html Yilmaz, K., Şen, V., Aktar, F., & Onder, C. (2021). Does Covid-19 in children have a milder course than influenza? The International Journal of Clinical Practice, 75(9), e14466. https://doi. org/10.1111/ijcp.14466

Chapter 3

The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

During the Covid-19 pandemic, people were constantly under the delusion of an exaggerated threat. When Sars-Cov-2 started circulating, mass hysteria was ignited and the irrational fear that everybody faces high likelihood of death was taken as gospel. The exaggerated IFRs that became known may have played a part in this, but they were not the only factors that contributed to mass hysteria. After all, it was not the first time—and certainly it won’t be the last—that false scientific projections take place. The public was inundated with warnings (primarily by the media) that serious disease and death are around the corner, and these reports did not respond to changes in the scientific developments. So even when science offered evidence that the actual risk of death is much lower than it was initially assumed or that schools were fully safe for reopening, the media coverage did not change its stance and mass hysteria did not recede. Governmental interventions, mainly lockdowns, exacerbated the effect by leading to a novel kind of mass hysteria, what I would like to call as prolonged mass hysteria.

3.1 Laying Out Mass Hysteria In mass hysteria, people start to believe that they will be exposed to something dangerous, such as a virus or a poison, because someone says so or because this fits their experience. Individuals from all walks of life can fall prey to mass hysteria, and thus, it is not a condition to be associated with pre-disposition towards psychological disturbances or with different educational and social backgrounds. The empirical evidence for such collective anxiety date back at least to Middle Ages (Starkey, 1949), and several episodes are to be found since then. One of the most astounding instances of mass hysteria in modern times was caused by a radio play named War of the Worlds, which described an attack from Martians. Many listeners thought that they were really under attack from Martians, and panic permeated the © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Karadimas, The Covid-19 Pandemic, Studies in Public Choice 42, https://doi.org/10.1007/978-3-031-24967-9_3

59

60

3  The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

atmosphere in the United States. Another example of mass hysteria was related to a Portuguese TV show in which the characters got infected by a life-threatening virus. After this, thousands of pupils reported symptoms similar to the ones the characters of the show suffered. As a result, many schools closed. However, the Portuguese National Institute for Medical Emergency showed that the virus did not exist and that the symptoms were caused by the anxiety and the panic that broke out. More recently, on the Emirates flight 203  in 2018, some passengers developed flu-like symptoms. When other passengers observed them, they started to feel sick, too. Panic broke out and the flight was quarantined when it arrived in New York. In fact, though, the investigation showed that only a few passengers had seasonal flu or the common cold (Bagus et al., 2021). Mass hysteria, which is often considered synonymous to terms like “epidemic hysteria” or “mass psychogenic illness,” can occur in several settings such as schools, public and private businesses, as well as small communities, and affects pretty much all ages (Boss, 1997). This is an umbrella term and can be used to describe a large amount of unexpected collective nonsensical behavior. A classic distinction in the literature is the one offered by Wessely (1987), who identifies two types of mass hysteria: “mass anxiety hysteria” and “mass motor hysteria.” The former occurs mainly among schoolchildren and involves sudden, acute anxiety, while the latter involves all age-groups, and apart from severe stress it also involves alterations in psychomotor activity. While it has been questioned whether this definition captures all cases of mass hysteria, since mass psychological and physiological abnormalities may overlap and sometimes co-occur (Ali-Gombe et al., 1997), it seems to be widely accepted among experts in the field that a false perception of a real or an imaginary threat can lead larger or smaller groups of people to suffering  an array of symptoms ranging from the ones expected by a physical disease (fever, fatigue, dizziness, etc.) to symptoms related to psychiatric illness such as convulsive movements or to make them being in a state of not well-grounded fear. Based on these observations, one could assume that mass hysteria affects those that are predisposed to some sort of psychological disturbance. However, it can affect everyone (Bartholomew & Wessely, 2002), and it does appear to affect most of the people in the setting in which a hysteria outbreak happens. Thus, while its symptoms are relevant to the ones induced by diseases, mass hysteria is not an illness, but rather a phenomenon that from time to time appears in societies. Perhaps mass hysteria exemplifies how blurred the distinction between who is mentally ill and who is not may be. As some argue, what is described as mental illness can be also conceptualized, at least in some cases, as having altered preferences or as having extreme preferences and thus the demarcation criterion for diagnosing mental illness is not always clear cut (Szasz, 1997; Caplan, 2006). One way to view mass hysteria thus could be to describe it as a state of affairs whereby groups of people suddenly and inexplicably alter their preferences and/or priorities. Scholars often draw parallels between the way germs spread and the way hysteria spreads (Phoon, 1982). Le Bon analyzes contagion outside the context of disease-­spread and focuses especially on how beliefs, sometimes odd ones, spread among the population. He argues that this powerful sort of transmission enforces

3.1  Laying Out Mass Hysteria

61

not only opinions but also feelings on other people (Le Bon, 1896/2002). Thus, when panic starts off, it is likely that many people will be overwhelmed by this feeling, at least in the setting in which the panic-outbreak happened (a school, a business et al.). Mass hysteria is aided and abetted by the development of novel patterns of behavior, the so-called emergent norms (Freedman, 1982; Van Ness & SummersEffler, 2016). So, when a group of people starts behaving in a certain way, this may be established as a norm, which could lead to an increase in the transmission levels of panic. When panic starts due to a gibberish belief and emergent norms ensue, these act as an amplifier of mass hysteria and likely spread it across the population very quickly. In my view, emergent norms can lead us from a state of local mass hysteria to widespread mass hysteria. Cases of local mass hysteria are confined within the setting in which the outbreak occurs. Episodes of mass hysteria induced by rumors of gas poison are examples of this sort. From March to April 1983, Palestinians reported symptoms such as headache, dizziness, and even fainting. Sixty-four people rushed to doctors after being under the delusion of having been poisoned with contaminated gas. After several tests, it was proved that there was no gassings and the symptoms rapidly went away. This and similar episodes occurred in schools and in some villages and did not spread outside these settings (Modan et al., 1983). On the contrary, cases of widespread mass hysteria occur when people develop emergent norms in a vain attempt to tame their fear. These norms spread hysteria farther afield and often tend to establish themselves as a normal pattern of behavior, thus making the hysteria a fixed state of affairs. Outbreaks of dancing mania that occurred in Middle Ages (1374, 1463, and 1518) can be described that way. It was held at that time that dancing was both the affliction and the cure of fear, and it was moreover considered as a way to save one’s soul. Thus, people were desperately dancing for months imploring priests to save their souls. As if that was not on its own hysterical enough, during the 1518 dancing outbreak, the authorities in Strasbourg made dancing compulsory, and people were mandated to dance day and night. This obviously led to a dramatic escalation of the epidemic, and it is estimated that about 400 people died (Waller, 2009). Panic-induced bouts of dancing cannot be explained solely in psychiatric terms, for it is unlikely that all people who danced unstoppably suddenly suffered simultaneously the same mental illness. Rather it seems that it was an emergent norm that had spread far and wide the belief that dancing saves souls. Another striking fact of these medieval episodes of mass hysteria, which, as we will see, finds application to the covid-related panic, is that the intervention of authorities made matters even worse by further magnifying the hysteria. Findings from psychology on risk-perception help us grasp the public’s reaction when mass hysteria is inflicted, especially when it comes to contemporary episodes of mass hysteria. Among others, biased media coverage, incomplete information, and personal experience lead to distorted risk-estimates (Slovic, 1987). Although it is difficult to identify one and only source as the leading cause of mass hysteria and even though the literature is far from being conclusive on this, it appears that these three strands of social life are, when become dominant, confluent factors that lead to mass hysteria. Misinformation gives distorted view of a subject matter, and it

62

3  The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

often leads to mistaken risk-estimates and induces not well-grounded fear. The dancing mania was certainly the result of an unnecessarily instilled fear in the population, and it was a corollary of having blurred vision of issues related to Christianity. As for the Covid-19 crisis, the false belief that all people face the same risk regardless of age was taken for granted and was barely questioned. Some inflated IFRs that appeared early on in the pandemic may have contributed to mass hysteria, though this on its own is not enough to ignite such collective fear because it is not uncommon to have mistaken scientific predictions in place, and it cannot be convincingly argued that false scientific projections are able to create mass hysteria. Neither the declaration of a pandemic by the WHO was enough to cause mass hysteria on its own for, as we saw, several times in the past a pandemic was declared, such as the H1N1 pandemic in 2009, and no mass hysteria ensued. Media coverage has certainly played a great part in establishing fear, and it turned out that in the digital age, it is easier for misinformation to spread than it was, for example, in the Middle Ages. Thus, when mainstream media across the globe reported pretty much the same piece of information, then most people in almost all countries took this to be a matter of fact. In general, it became nearly commonplace in the media that when it comes to Sars-Cov-2, all is doom and gloom. Researchers have shown that negative news on Covid-19 far outnumbers positive news. For example, stories of increasing Covid-19 cases outnumber stories of decreasing cases by a factor of 5.5 even during periods when new cases are declining (Sacerdote et al., 2020). Overall, bad news is considered as having stronger impact than good news (Baumeister et  al., 2001), and hence, their role in the development of mass hysteria is very important. Horrific stories of people collapsing in the streets and dying as a result of Covid-19 disease while members of medical staff wearing protective equipment were approaching them were broadcasted by media across the globe (Guardian, 2020a; Smith, 2020), and even though these videos were debunked because there is no such thing as a sudden coronavirus death, they definitely conveyed the message that a health disaster is looming. Moreover, despite the fact that it was clear from the early days of the pandemic that Sars-Cov-2 poses a threat mainly to old individuals, people ended up believing that half of covid cases will be hospitalized. As it was noted in the previous chapter, the hospitalization rates for Covid-19 are comparable to the hospitalization rates for influenza and nowhere near to 50% as many panicked people believed or even to 20%, as some pundits falsely claimed early on in the pandemic. Even more astonishingly, people failed not only to make an approximately accurate age-adjusted estimate of the risk the virus poses, but they totally missed it, and the majority of the population was deeply convinced that all people face at about the same risk of dying from Covid-19. Some polls showed jaw-­ dropping results. People believed, for example, that someone who is 24 years old has an 8% likelihood of death (Rothwell & Desai, 2020), whereas in fact the IFR for this age-group is lower than 0.1%, i.e., lower than the risk that influenza poses to that age-group. Part of the misinformation campaign was moreover the overplayed issue of asymptomatic transmission which resulted in people believing that healthy individuals can transmit a very deadly pathogen, despite the fact that, as we discussed in chapter 2, the relevant mistaken report was quickly criticized and rejected

3.1  Laying Out Mass Hysteria

63

by the scientific community. It is clear that misinformation due to biased media coverage ignited mass hysteria in the early days of the pandemic and was an important ingredient that helped keeping mass hysteria on going for a very long time. By contradicting data and by taking at face value that all people face high risk of dying and that healthy people can kill other healthy people just by talking or touching them, the world plunged, within a few weeks, into unprecedented scales of panic. Emergent norms quickly showed up and amplified mass hysteria. These emergent norms included behaviors that treated healthy people as infected, and thus, people hoped that by adopting the novel norms, they could be saved from the supposed plague. Avoidance of handshakes was perhaps the most significant emergent norm and the one promoted very much by the media (Guardian, 2020b).1 Of course, there is no scientific basis to this attitude, and it plays no role into transmission because first, we saw that asymptomatic transmission is hardly a fact, and second, even if we assume that asymptomatic transmission is common, Sars-Cov-2 is not transmitted through fomites, and thus, someone who has touched a contaminated object will not pass an active virus fragment on to someone else through handshake. Another emergent norm was that people were staying apart from each other and the mantra “stay six feet apart from each other” prevailed. People kept distance from each other not only indoors but also outdoors as well. Since there is no reason to consider healthy people as infected carriers of a very deadly virus, then staying “six feet apart” from each other is no more than a panic-induced emergent norm. Staying distant from each other when indoors deviates from what we considered normal until early 2020, but when this occurs outdoors, it is even more nonsensical because transmission of Sars-Cov-2 happens almost exclusively indoors (Gov. UK, 2020). A third instance of mass hysteria, which occurred a few months after the initial outbreak of panic, is that people were wearing masks even in places where they were not asked to do so. We saw that the evidence on masks suggests that they do nothing to stop transmission, so wearing them when you are not coerced by the law is not a science-based attitude but rather a reaction to the panic that permeated. Thus, if one argues that people did their own research and concluded that masks should be worn even when it is not required to wear one, it could be replied that this barely holds for the results of such a research ought to have made them question the frenzy over masks altogether, not to passionately wear them all day long. So either they did research and they came up with the exact opposite of a proper inference or they did not do research and something else explains this behavior and it appears that irrational fear does. It seems therefore that these three emergent norms were the aide that established mass hysteria and spread it very quickly throughout the population. The first two are a direct consequence of the initial stages of mass hysteria, while the third is a corollary of what governmental interventions seem to have caused in this crisis, namely prolonged mass hysteria.

 Some coined the term “footshakes” and called for replacing handshakes with footshakes in order to avoid the virus (McKeever, 2020). 1

64

3  The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

3.2 Mass Hysteria Intensified and Drew Out A theory of mass hysteria in the context of the Covid-19 crisis was developed first by Bagus et al. (2021) who examine how governmental interventions can generate mass hysteria and how the size of the state can lead to different levels of mass hysteria. While I think that their work is pioneering and that it squares very well with what has happened, I beg to differ in some respects. They claim that mass hysteria started off because of mitigation measures and of a centralized approach to the virus, and that in the absence of a welfare state, such a collective panic would have not been so dominant. In theory that is possible, but during the covid crisis, the measures came in when panic was already prevalent. At first glance, the issue of which came first, the panic or the lockdowns, is the similar to the puzzle of “which came first? The hen or the egg?” However, the media coverage on the supposed plague was incessant and started off with the Wuhan outbreak and the stories of sudden covid deaths. Thus, by March 2020, the media had convinced the majority of the population that they will be exposed to a serious death-threat while the emergent norms had spread the hysteria in almost every nook and cranny of earth, and thus in early March 2020, when no country but China had locked down, mass hysteria was dominant in almost all countries in the world. People were ready to self-isolate even if they were not enforced by the government to do so; as we saw, avoidance of handshakes became an established pattern of behavior well before lockdowns were in place; buying unreasonably high stocks toilet papers in early March 2020 is another good indicator of mass hysteria prior to lockdowns (Stratton, 2021). More importantly, we have data indicating that even in non-lockdown countries, such as Sweden, people were also infected by mass hysteria during the early days of the pandemic, and thus it was not the case that in nonlockdown countries people were not panicked. On the contrary, this was a setting whereby panicked people did their best to avoid a—what was then conceived as—serious death-threat, and if a government acts in a way that shows that it shares the same concerns, it greatly maximizes its own utility. In view of this, lockdowns came in to meet the demands of an already panicked public, and it is not that governments ignited hysteria in the first place. What Bagus et al. (2021) got absolutely right was that sweeping mitigation measures do play a role in mass hysteria by implying that a health calamity is on the way. Thus, by implementing lockdowns, which is a quarantine not only for the infected but also for the healthy ones, and by enforcing mandatory mask wearing to everyone (in some places masks were mandated even outdoors) signals that the risk of dying is really huge, by vindicating that this risk is nearly the same across age-­ groups and that in the absence of such measures an untold plight is to be expected. Thus, lockdowns and the like made mass hysteria orders of magnitude worse than it already was, and we were led to prolonged mass hysteria. Prolonged mass hysteria made emergent norms such as avoidance of handshakes and mask wearing almost the new normal, and people were expecting from the governments to keep imposing strict measures in order to protect them. Thus, the behaviors of people wearing masks when it was not required by the law (an emergent norm that followed a few

3.2  Mass Hysteria Intensified and Drew Out

65

months after panic broke out) appear to be explained best in the context of prolonged mass hysteria, where fear was deeply instilled in the population. Should politicians made no intervention, mass hysteria would have faded away and the emergent norms would not be prevalent for a very long time. This is not because people would update their views on pandemic policy based on new evidence, but because the absence of lockdowns gives them the chance to gather information about the actual virus-related risk of everyday activities. People often hold beliefs that do not follow from the available evidence or beliefs that scientists rate them as wrong, but since acting on false beliefs usually bears a personal cost (often in terms of material well-being), people estimate tradeoffs between sticking to their own mistaken beliefs and their everyday interests. So, for example, they may believe that free trade is bad for the economy but ignore this belief during their everyday transactions. However, there is a setting in which the cost of practicing one’s own pre-existing biases is effectively zero, and this is during the voting procedure. As public choice theorists argue, people have no strong incentive to revise their biases on public policy issues when new evidence appears because the probability of one single vote to influence the election outcome is nearly zero and the cost of voting based on mistaken beliefs is also zero. Thus, people can collectively choose bad policies by making an individually rational choice during voting. On the contrary, they do have incentives to take new evidence into account when making personal decisions in order to minimize the possible cost they will likely suffer otherwise (Caplan, 2001, 2007). During the Covid-19 pandemic, as mass hysteria prevailed, people falsely believed that all age-groups face high risk of death and that asymptomatic transmission is common. In non-lockdown settings, people were initially following the emergent norms and were more cautious in their everyday lives, but since no lockdowns were implemented, they realized that staying voluntarily at home for a very long time bears a remarkable personal cost and makes them worse off. They thus decided to estimate personal tradeoffs without substantially revising their false, mass-hysteria-induced, beliefs. So, for example, not going out with friends bears the cost of loneliness and meeting them at a restaurant carries the risk of catching the virus. Dining out with friends and returning home healthy makes a person realize that one is better off going to a restaurant with friends than staying at home. Catching the virus and experiencing only mild symptoms has the same effect. Similar tradeoffs on several aspects of everyday life estimated by many people at an individual level led to a gradual declining of mass hysteria in countries that did not lockdown, but it does not follow that people were less alarmist about the virus; they simply chose material interests over beliefs because they could not bear the cost of acting on their own beliefs. Data from Sweden show that although the Swedish people were not immune to mass hysteria, the absence of lockdowns made life return to almost normality rather quickly, despite the fact that Sars-Cov-2 continued to circulate. From mid-March 2020 to early April 2020, people in Sweden avoided public transportation and visiting shops, but since schools and businesses remained open, people returned to their old habits, and mass hysteria was almost entirely defeated (OWD, 2021). So a possible way out of mass hysteria, at least with respect to the Covid-19 crisis, is to press politicians to stop intervening and let people

66

3  The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

estimate their own tradeoffs. By implementing policies that magnify an already exaggerated threat—and given that the media continue not highlight positive news on covid—prolonged mass hysteria will likely continue to be the norm (Karadimas, 2022). Figure 3.1 summarizes the case.

Misinformation Breaks Out

Panic Starts Off

(Media give distorted view of risk)

No Emergent Norms: Local Mass Hysteria

Mass Hysteria Confined Emergent Norms Ensue

Widespread Mass Hysteria No Governmental Action Governmental Intervention

(Personal tradeoff estimation)

(Lockdowns)

Mass Hysteria Declines Mass Hysteria Magnified

Prolonged Mass Hysteria Prevails Fig. 3.1  Pathways to (prolonged) mass hysteria

3.3  Mass Hysteria and Social Desirability Bias: Panic Institutionalized

67

This is an overview of mass hysteria as it developed during the Covid-19 pandemic, and while it may apply to other episodes as well, I do not contend that it necessarily captures all possible cases of mass hysteria.

3.3 Mass Hysteria and Social Desirability Bias: Panic Institutionalized People often need to be on a par with socially desirable etiquettes and patterns of behavior, and so they tend to appear more altruistic than they actually are by denying behaviors that would be considered socially undesirable (Zerbe & Paulhus, 1987). This seems to be very common when surveys are conducted and people respond to questions on sensitive topics, such as sexual behavior or illegal actions, e.g., driving a car while having exceeded the permitted alcohol limits, and they give answers that other people would consider them as desirable (Krumpal, 2011). A typical example of attitudes distorted by social desirability bias is pro-­environmental answers. People say in surveys, for example, that they recycle on a daily basis even though that may not be true (Kormos & Gifford, 2014). Psychologists and social scientists examine social desirability bias primarily within the context of investigative research and  explore how socially desirable responses could confound their results. As some of them have pointed out, the fact that such researches are anonymous often makes social desirability bias a minor issue (Fox & Schwartz, 2002), but the jury is still out on this (Vesely & Klockner, 2020). However, social desirability bias is obviously not limited to participants’ responses in surveys and finds applications to real life, too (Chung & Monroe, 2003). Speaking of environmentalism for instance, people do not want to be thrown epithets on them and be labeled as “climate deniers,” and thus they barely openly oppose climate alarmism, despite the fact that they may actually not enthuse over it. In a similar vein, presenting oneself as cautious about health issues is indeed socially desirable. When a health issue is discussed under conditions of mass hysteria, as is the case with Covid-19, social desirability bias appears in a subtle way that strengthens mass hysteria and acts in favor of it and of its emergent norms. Thus even though the vast majority of the population was indeed really panicked regardless of social and educational background, many people that either had access to the available data or who had simply realized that the danger of Covid-19 is overhyped, not only they did not speak out but they also extolled the mitigation measures, and, when lockdowns were lifted, emphasized that every single individual and every single institution should abide by lockdown-like protocols such as mask mandates and vaccine passports. Of course we cannot be sure who was knowledgeable and who was not, but when professors of medicine appear in the media and insist that masks are effective or that vaccine passports could help control the disease despite the fact that it was abundantly clear that masks do not work and that vaccines do not stop transmission, they were either blatantly ignorant, which is hardly the case, or

68

3  The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

they were craving for social desirability. The very few professors who publicly challenged lockdowns and offered an alternative policy proposal as well as an age-­ stratified analysis of the risk of dying from the virus or those who accurately summarized the data on masks were libeled and demonized as obviously stupid or venal persons. The authors of the Great Barrington Declaration are profound examples of this sort. They were vilified by mainstream media across the globe for simply saying publicly what was scientifically accurate. The pervasive panic left no room for a more reasonable approach to the virus, and thus many experts did not consider that there is fertile ground for scientific accuracy, and thus they either refrained from speaking to the media or, when they did so, their public messages were on a par with the mainstream view. They thus parroted what most people believed: this is a very deadly pandemic, lockdowns are necessary and the government is right to enforce strict restrictions and protocols everywhere. If more professors of medicine have questioned the prevailing view, then there could have been a chance that mass hysteria would have receded and that the governments would have been more cautious in implementing such restrictions. But since most of them conveyed a socially desirable message, then mass hysteria was consolidated and it became even more difficult to calmly explain that Covid-19 is not the new plague and that mitigation measures do very few, if anything at all, to stop a respiratory pathogen. This seems to explain two things: first, why the lockdown orthodoxy faced almost no pushback and was considered as of unmatched wisdom for a very long time not only among the general public but also among the so-called “elites,” such as members of the teaching staff in universities, judges, and other cultivated people. Second, it explains why panic was deeply fixed and institutions did not even dare to consider an alternative. Although most of the “elites” seem to have been genuinely overwhelmed by mass hysteria and failed to rationally assess new information, we can assume that among them there were some non-experts, i.e., people with no medical training, who had realized that the virus is not as deadly as it was thought that it would be in the first place and that governments need to substantially revise their strategy. These people were also hesitant into speaking against the narrative because they could be called out as “deniers” who do not follow science and who have no problem to lead society to a virus-induced massacre. But even those who did not pay that much attention to the scientific part of the pandemic failed to raise ethical concerns that, at least until early 2020, would have been considered legitimate. For example, lockdowns are an unprecedented restriction to human movement, and this on its own ought to have been enough to raise eye-brows. Furthermore, when lockdowns were lifted for a while and even more emphatically when vaccines became widely available, most institutions, including Universities, Courts, as well as religious institutions, all of them enthusiastically implemented protocols, such as mask mandates and showing proof of vaccination to enter a building, that not only they were not backed up by science but they are furthermore ethically concerning. Even though this is not the place to pick up on ethical issues, it appears that asking someone to divulge their vaccination status is a flagrant breach of medical privacy and imposing such a measure would be problematic even if covid’s fatality rate was equal to that of smallpox’s, but it becomes even more objectionable when taking

References

69

into account that it poses a high risk only to a certain chunk of the population. However, questioning the validity of lockdowns or of the protocols that prolong the hysteria was considered unethical in its own right, and almost no one wants to present themselves as someone who does not care for public health and who puts their fellows at risk by casting doubt on the necessity of the measures taken. The emergent norms therefore were further established and observing them became a typical way to present oneself as socially desirable. Indeed, they seem to have dominated every aspect of daily life: Avoiding handshakes means you care about not infecting others despite the fact that you are healthy and that handshakes do not transmit Sars-Cov-2. On the contrary, shaking hands means you are an idiot who puts themselves and others to an untold threat through a reckless sense of bravado. Wearing a mask is also virtue signaling and not wearing one indicates that one does not take the pandemic seriously enough, and thus they do not follow science and hence they should be vilified. Thus, as if it were not enough that the majority of the population was infected by mass hysteria, the few people who did not buy the fear mongering either remained largely silent or even praised lockdowns and followed the emergent norms, for that was the socially desirable thing to do. In brief, social desirability bias solidified mass hysteria.

References Ali-Gombe, A., Guthrie, E., & Dermott, M. (1997). Mass hysteria: One syndrome or two? British Journal of Psychiatry, 170, 387–388. Bagus, P., Peña-Ramos, J. A., & Sánchez-Bayón, A. (2021). Covid-19 and the political economy of mass hysteria. International Journal of Environmental Research and Public Health, 18, 1376. https://doi.org/10.3390/ijerph18041376 Bartholomew, R., & Wessely, S. (2002). Protean nature of mass psychogenic illness. British Journal of Psychiatry, 180, 300–306. Baumeister, R.  F., Bratslavsky, E., Finkenauer, C., & Vohs, K.  D. (2001). Bad is stronger than good. Review of General Psychology, 5(4), 323–370. Boss, L. P. (1997). Epidemic hysteria: A review of the published literature. Epidemiologic Reviews, 19(2), 233–243. Caplan, B. (2001). Rational irrationality and the microfoundations of political failure. Public Choice, 107(3/4), 311–331. Caplan, B. (2006). The economics of Szaz: Preferences, constraints and mental illness. Rationality and Society, 18(3), 333–366. https://doi.org/10.1177/1043463106066377 Caplan, B. (2007). The myth of the rational voter: Why democracies choose bad policies. Princeton University Press. Chung, J., & Monroe, G. S. (2003). Exploring social desirability bias. Journal of Business Ethics, 44, 291–302. Fox, S., & Schwartz, D. (2002). Social desirability and controllability in computerized and paper-­ and-­pencil personality questionnaires. Computers in Human Behavior, 18, 389–410. Freedman, J.  L. (1982). Theories of contagion as they relate to mass psychogenic illness. In M.  Colligan, J.  Pennebaker, & L.  Murphy (Eds.), Mass psychogenic illness (pp.  171–182). Routledge. Gov. UK. (2020). PHE: Factors contributing to risk of SARS-CoV2 transmission in various settings, 26 November 2020. https://www.gov.uk/government/publications/phe-­factors-­contributing-­to-­ risk-­of-­sars-­cov2-­transmission-­in-­various-­settings-­26-­november-­2020. Accessed 18 Dec 2020.

70

3  The World Stampeded: From Mass Hysteria to Prolonged Mass Hysteria

Guardian, The. (2020a). A man lies dead in the street: The image that captures the Wuhan coronavirus crisis. https://www.theguardian.com/world/2020a/jan/31/a-­man-­lies-­dead-­in-­the-­street-­ the-­image-­that-­captures-­the-­wuhan-­coronavirus-­crisis. Accessed January 31 2020. Guardian, The. (2020b). The end of handshake: Saying hello during the coronavirus outbreak. https://www.theguardian.com/world/2020b/mar/03/the-­end-­of-­the-­handshake-­saying-­hello-­ during-­the-­coronavirus-­outbreak. Accessed 3 Mar 2020. Karadimas, P. (2022). Covid-19, public policy and public choice theory. The Independent Review; A Journal of Political Economy. 27(2), 273–302. Kormos, C., & Gifford, R. (2014). The validity of self-report measures of proenvironmental behavior: A meta-analytic review. Journal of Environmental Psychology, 40, 359–371. Krumpal, I. (2011). Determinants of social desirability bias in sensitive surveys: A literature review. Quality and Quantity, 47, 2025–2047. https://doi.org/10.1007/s11135-­011-­9640-­9 Le Bon, G. (2002). The crowd: A study of the popular mind. Loki’s Publishing. McKeever, V. (2020). The coronavirus is stopping the handshake… but the footshake taking its place. CNBC. https://www.cnbc.com/2020/03/05/the-­coronavirus-­is-­seeing-­the-­footshake-­ replace-­the-­handshake.html. Accessed 5 Mar 2020. Modan, P., Tirosh, M., Weissenberg, E., et  al. (1983). The Arjenyattah epidemic. The Lancet, 322(8365–8366), 1437–1512. https://doi.org/10.1016/S0140-­6736(83)90815-­2 OWD. (2021). Coronavirus pandemic data explorer. https://ourworldindata.org/grapher/visitors-­ transit-­covid?tab=chart. Accessed 18 Apr 2021. Phoon, W. H. (1982). Outbreaks of mass hysteria at workplaces in Singapore: Some patterns and modes of presentation. In M. Colligan, J. Pennebaker, & L. Murphy (Eds.), Mass psychogenic illness (pp. 21–32). Routledge. Rothwell, J., & Desai, S. (2020). How misinformation is distorting COVID policies and behaviors. Brookings. https://www.brookings.edu/research/how-­misinformation-­is-­distorting-­covid-­ policies-­and-­behaviors/. Accessed 22 Dec 2020. Sacerdote, B., Sehgal, R., & Cook, M. (2020). Why is all Covid-19 news bad news? (NBER working paper no. 28110). National Bureau of Economic Research. Slovic, P. (1987). Perception of risk. Science, 236, 280–285. Smith, O. (2020). Coronavirus horror. Express. https://www.express.co.uk/news/world/1232814/ Coronavirus-­horror-­China-­virus-­Wuhan-­zombies-­epidemic-­video. Accessed 24 Jan 2020. Starkey, M. L. (1949). The devil in Massachusetts, a modern inquiry into the Salem witch trials. A. A. Knopf. Stratton, J. (2021). Coronavirus, the great toilet paper panic and civilization. Thesis Eleven. https:// doi.org/10.1177/07255136211033167. Accessed 20 July 2021. Szasz, T. (1997). Insanity: The idea and its consequences. Syracuse University Press. Van Ness, J., & Summers-Effler, E. (2016). Reimagining collective behavior. In S. Abrutyn (Ed.), Handbook of contemporary sociological theory (pp. 527–546). Springer. Vesely, S., & Klockner, C. (2020). Social desirability in environmental psychology research: Three meta-analyses. Frontiers in Psychology, 11, 1395. https://doi.org/10.3389/fpsyg.2020.01395 Waller, J. (2009). The art of medicine: A forgotten plague: Making sense of dancing mania. The Lancet, 373, 624–625. Wessely, S. (1987). Mass hysteria: Two syndromes? Psychological Medicine, 17, 109–120. Zerbe, W. J., & Paulhus, D. L. (1987). Socially desirable responding in organizational behavior: A reconception. Academy of Management Journal, 12(2), 250–264.

Chapter 4

Tradeoffs and Knock-On Effects

Apart from failing to tackle the pandemic, lockdowns caused serious side effects. In fact, policy makers failed to take into account the fundamental tradeoff between the lives that are threatened by the mitigation measures and the lives that are threatened by the virus and claimed that the tradeoff was between the lives that are threatened by the virus vs. the money that some will lose and the fewer chances for entertainment and social activity. The money lost and the restricted socialization were considered as innocuous outcomes, and they could have been so if they were not linked to reduced lifespan. The mitigation measures in place, especially the economic shutdown as well as the school closures, wreaked havoc in society and reduced the longevity of people that would have hardly died from the virus. The years of life lost (YLL) and the hazard ratios (HRs) help us estimate the consequences of this policy.

4.1 The Wrong Dilemma A cost-benefit analysis of the effects of a policy is always an important part of both scientific practice and of policy making. Cost-benefit estimates typically sum the potential benefits from a scheduled policy (or a business plan) and compare this with possible weaknesses and/or expected losses. When examining policy-making issues, especially issues related to public health, economists usually estimate the value of a life by assigning a dollar value to individual’s life and by figuring out how many dollars one is willing to pay in order to reduce risk. This measure is well-­ known as “value of statistical life” (VSL), which was first developed by Thomas Schelling (1968). It represents the rate of substitution between wealth and mortality risks, and it estimates the benefits of lives saved in monetary terms. Anderson and Treich give a simple example which shows how the concept is used. Suppose that in a city of 100,000 citizens, approximately 5 citizens die each year on its roads. The authorities now set out a task to reduce mortality in these streets from 5 to 2 deaths © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Karadimas, The Covid-19 Pandemic, Studies in Public Choice 42, https://doi.org/10.1007/978-3-031-24967-9_4

71

72

4  Tradeoffs and Knock-On Effects

per year and assume that each citizen is willing to pay $150 each year to benefit from this reduction in mortality. $15million could be collected to save 3 statistical lives, and it thus follows that the corresponding VLS is $5million per life—that is the value of a statistical life in this example (Anderson & Treich, 2011). This example describes a project that passes the cost/benefit test because each life is valued as $5million, while the cost per person is $150. The VSL measure is commonly used in several cost/benefit analyses related to the Covid-19 pandemic. The typical estimate from most scholars includes a VLS estimate of the lives saved by mitigation measures (primarily lockdowns), and this value is compared to the fall in GDP that occurs. Most of these studies make far-­ fetched assumptions about the efficacy of lockdowns, the lethality of the disease, and the transmission rates and, based on these assumptions, conclude that lockdowns save many lives at some cost in GDP growth. By and large, they postulate the three mistaken theses of pro-lockdown rhetoric described above, combined with the undocumented hypothesis that the more measures taken the more lives are saved, and they come up with a VSL that is likely higher than the parameter value of the loss in GDP, and thus the conclusion they draw is that lockdowns are cost-effective. This type of studies comes primarily from the early days of the pandemic and the authors are mainly economists, which perhaps explains why they took at face value some mistaken projections regarding the disease severity that became known early on in the pandemic and/or that they were overwhelmed by mass hysteria and believed that this is an unprecedented health crisis. An account offered by Thunström et al. (2020) belongs to studies of this sort. They come up with a VSL of $12.4 trillion of the 1.24 million lives supposedly saved due to lockdowns and that the cost includes a 6.2% fall in the GDP which equals $7.2 trillion. If that is the case, then shutting down the economy is an excellent policy. Studies like this one, apart from making extraordinary assumptions about the ability of lockdowns to reduce death-­ rates and the risk Covid-19 poses to the population, they moreover misuse the VSL estimate by taking it to be constant across all age-groups. However, that is not how the VSL is usually used. Assuming that the VSL is constant is like claiming that individuals are indifferent between living one more day and 80 more years, an estimate that further distorts the results (Hammitt, 2020). A more reasonable way to conduct such cost/benefit studies is to posit that the VSL decreases as age increases, which could lead to more sound results (Robinson et al., 2020). Indeed, someone who is 80 years old and has serious comorbidities is unlikely to pay the same amount of money as someone who is 20 years old and cost/benefit analyses need to take this into account. However, even if that approach could lead to more sensible conclusions, these studies keep comparing lives to money lost, to jobs lost, or to reduced socialization. First off, this creates the problem of how to add up all these costs into one single distribution and compare them against the (assumed) benefits. To avoid this disturbance Bryan Caplan proposed another measure which slightly deviates from the standard VSL approach. Instead of estimating how much one is willing to pay to avoid death from Covid-19, he calls for estimating how much an individual is willing to pay to avoid living under lockdown. Thus, the total costs of lockdown policies

4.1  The Wrong Dilemma

73

as they were perceived by each person would be the cost utility which would stand against the lives saved by the lockdowns. Caplan puts it in terms of time rather than money and sets value X as the number of months one is willing to sacrifice to avoid living under lockdown. Since it is highly unlikely that all people would be willing to sacrifice the same number of months, Caplan reaches an average of 2 months per person. Building on Caplan’s approach, Allen applies the same average value (X = 2) to estimate costs and benefits of the Canadian lockdown. He also makes improbable assumptions about the efficacy of lockdowns, and he takes for granted many of the failed predictions of the pandemic modeling. Thus, he postulates that without a lockdown death-rates would have been 20–50% higher than they were under lockdown. This is intentional, for the author concludes that even under such crazy assumptions about the possible death-rates, lockdowns do not pass the cost/ benefit test (Allen, 2021). Although informative and useful, the discussion on cost/benefit estimates and mainly of those studies using the VSL measure as a calculation tool of the costs and benefits induces the wrong tradeoff. It measures health against other things in life people value, such as job, entertainment social relationships, and so forth. Even if we use it, as many economists did, to estimate the tradeoff between lives and a fall in GDP, it again creates a misleading framework that does not help very much to explore the actual damage done by the mitigation measures, for it does not take into consideration the strong correlation between fall in GDP and reduced lifespan or the life-threatening consequences of reduced income or reduced social relationships. If that is taken into account, then the tradeoff instantly changes and it no longer compares money against money, or lives saved against money spent, but lives against lives. While governments and lockdown supporters never confronted themselves with a task of developing a cost/benefit analysis of their pandemic policies, the way early cost/benefit accounts were presented was on a par with what governments suggested, i.e., that the tradeoff is between the health and the economy, and this generated a consensus that underscored the (then assumed) benefits of a lockdown and undermined the possible side effects. This is absurd because the correlation between recession and reduced longevity was not new in economics, and in my view, the fact that many economists failed to take into account such knowledge was, as I have noted, the result of mass hysteria which made many people not only from the general public but also among the scientific domain and the Academia to firmly believe that Covid-19 poses a serious threat for all people, and thus unemployment is a minimal price to pay in exchange for avoiding the disease and staying alive.1

 Perhaps social desirability bias could be applied one more time to explain the fact that economists took for granted that the tradeoff is lives vs money. Indeed, if one attempted to object that this tradeoff is not compatible with some well-established knowledge from economics and social sciences they could be libeled and so they tried to show that lockdowns do not worth their cost by postulating premises similar to the three false ones discussed above so that their proposal would appear socially-friendly. Of course, genuine misunderstanding of the situation as a result of mass hysteria also explains their stance. 1

74

4  Tradeoffs and Knock-On Effects

Some economists seem to have been aware of this correlation and have tried to take this into account by using another method which is now popular among economists, that of QUALYs (quality-adjusted life years). It measures the health outcomes as they are experienced by the individual. QUALYS measures range from 0 (death) to 1 (perfect health). One year lived in perfect health equates to one QUALY; one year lived in nearly perfect health (in case, for example, a professional footballer cannot participate in matches due to an injury, but other than that is healthy), could equate to, say, 0.8 and so on (Whitehead & Ali, 2010). To calculate the precise number of QUALYs, one can multiply the utility value associated with a given state of health by the years lived in that state (Drummond et al., 1997; Prieto & Sacristan, 2003). In the footballer’s example, his QUALYs are 0.8 (0.8 × 1). While this measure encompasses the quality and quantity of the years lived and thus it is wider in scope than the VSL, for it leads to conclusions about possible life-threatening interventions, its main feature is that its outcomes are largely based on how individuals perceive health conditions, which creates problems when applied to the Covid-19 pandemic for it seems that it does not capture a key aspect of the story: that people were in a state of mass hysteria and had distorted risk estimates about the virus and about the interventions that would suppress the disease. As we have seen, they overestimated the risk from the virus and underestimated the risk from the mitigation measures. So it is not clear that people were aware about the damage inflicted by the lockdowns; in fact, the opposite seems to had been the case, i.e., people were under the delusion that mitigation measures saved them from certain death, and thus they could value living under lockdown as a life of nearly excellent quality, which could obviously lead to distorted conclusions when QUALYs are applied to calculate costs and benefits of lockdowns. Miles et al. (2020) use QUALYs measure to estimate the costs and benefits of the lockdown in the UK after the first pandemic wave. In short, they tried to calculate the QUALYs lost for each covid death, how many lives lockdowns saved and make comparisons with the cost in GDP. Thus, we see that the tradeoff here is, one more time, between lives lost due to the virus and a cost in GDP. The underlying assumptions they use on the efficacy of lockdowns are also out of touch, albeit intentionally, for they too take the predictions of the Imperial modeling to be close enough to reality, and while they recognize that 500,000 deaths in the UK in the absence of a lockdown was a pessimistic scenario, they, rather arbitrarily, postulate that the lockdown in the UK during the first wave saved “at least” 20,000 lives. However, even in such a setting, a lockdown fails the cost/benefit test. The authors conclude that the lowest estimate for lockdown costs incurred was 40% higher than the highest benefits from avoiding the worst mortality case scenario and argue for abandoning lockdown strategies (Miles et  al., 2020). We see therefore that even under such assumptions and given the fact that people were under the sway of mass hysteria during the Covid-19 pandemic (which creates problems for the QUALYs measure), lockdowns do not pass the test. So while I do not deny that using VSL or QUALYs measures could be a useful way to make cost/benefit estimates, I think there are some inherent problems with these two methods so that when they are applied to the covid crisis they bring to the

4.2  Estimating Tradeoffs: Established Knowledge as Guiding Principle

75

fore the mistaken dilemma. Mistaken dilemma is the one that makes comparisons that are at odds with the available theoretical and empirical knowledge. In other words, these comparisons are not the result of a thorough evidence-gathering process and thus they can hardly lead to a well-informed decision-making process. Even though it is still possible that one can use VSL and QUALYS and conclude that lockdowns are not an optimal policy, a lockdown advocate could reply that a loss in GDP is a harmless damage in comparison to lives saved. Lockdowns are very costly, and thus even if that was the case, lockdowners would still have a difficult task to accomplish, but at least they could have been right that the dilemma is between lives saved and money lost which is a setting that could perhaps strengthen a bit the pro-lockdown argument and the public-interest explanation. Of course, in view of the prolonged mass hysteria and the exaggerated claims about the threat covid could pose, it was clear that the public was convinced that this was the tradeoff. Indeed, if one was asked to answer the question “Do you prefer to stay unemployed and/or to see your income reduced or do you prefer to be exposed to a virus that is likely to kill you or at least to require you to be admitted to an ICU bed and even if you survive it is as likely as not that you will suffer poly-organic failure due to long-covid which suggests that you will never recover fully?” Nearly no one could reply that they prefer exposure to such a virus and since humanity was in a state of prolonged mass hysteria the consensus was established: unemployment and losses in GDP are not even conceived as costs. Thus, VSL and QUALYs measures do not make a strong case against lockdowns for they calculate costs in lives against costs in terms of GDP fall,2 and they are thus laid open to the sort of criticism just described. On the contrary, the criterion I am proposing takes into account the established knowledge and comes up with the tradeoff that is compatible with what is long-known, i.e., that we have lives that are threatened by the virus (or whatever germ is circulating) and lives that are threatened by the fall in GDP. While I focus here on the covid crisis, this approach could be applied to other issues when discussing well-being and when policy-making is in progress.

4.2 Estimating Tradeoffs: Established Knowledge as Guiding Principle John Broome is mindful that we are many times confronted to choosing between lives and thus recognizes that economists need to find a way to frame this so that decision makers will have a construction which will guide them in such cases. In his  While QUALYs allow for the possibility that quality of lives is taken into account, which suggests that if the inputs are altered then there is perhaps a pathway to estimate the side effects in terms of reduced quality of life, in my view this hardly applies to circumstances whereby mass hysteria governs the population’s attitude for, as we have seen, people make distorted risk-estimates (hence, they are not sober-minded when estimating “years in perfect health”) and thus are mostly unaware of the damage inflicted on them. 2

76

4  Tradeoffs and Knock-On Effects

book Weighing Lives, he undertakes the task to offer such a theory. Broome’s theory amounts to measuring well-being at an individual level and then at a group level. A person’s lifetime well-being is the sum of the values of all the so-called “temporal well-beings,” i.e., of instances of well-being throughout their life. As for the value distribution of a population’s well-being, Broome coins the term “neutral level for adding a life,” which is a level of a population well-being that is neither better nor worse to add a life in this population. He then suggests that the value distribution of the well-being of a population is the sum of the measures of individuals’ lifetime well-being each of whom is above the neutral level for adding a life. This is what Broome calls the “intergraded standardized total principle” (Broome, 2008). Broome is right that estimating individuals’ well-being at a particular time is an important task and his theory tries to offer a way to conceptualize how this can be done in advance, i.e., how we can possibly have one distribution that estimates individual and population well-being so that it will be able to guide us under any possible circumstances. Broome does not lack intellectual honesty and highlights the missing parts of his own theory. Thus, while he offers a theoretical framework regarding how well-being is to be estimated, he recognizes that empirical input is needed so that it can be determined how well off a person is at a particular time and he admits that this is a complicated project (Broome, 2008, 260). It is indeed not an easy task to estimate a person’s or a population’s well-being, but it seems to me that we can have a pathway to estimate how to make individuals as less worse off as possible which becomes of paramount importance when a novel problem appears and decision-making is to take place. The old theory of opportunity cost suggests that, even under the best possible circumstances there is hardly a zero-cost life (Knight, 1921). Thus, instead of trying to find a principle that will set the standards for well-being we should consider how to minimize damage. To do so, it is not enough to measure costs and benefits but, as we saw, we need to be able to consider the proper tradeoff which is far from clear that it is obvious that we will be able to do so, as it became clear during the Covid-19 pandemic. It appears that the best way to successfully achieve this two-pronged strategy is by taking into account long-established knowledge from the social sciences, to consider it alongside novel evidence that could appear, and then to take stock of the kind of decisions that are to be taken. In principle, this is, at least by and large, what public-interest theorists contend that politicians do but as it became increasingly clear during the covid-­ crisis, not only policy makers can fail to take it into account, but this is also possible among the scientific community. By bringing to the fore what I like to call as “the proper tradeoff” which is, as already stated, the tradeoff that is compatible with sound theoretical and empirical knowledge that is on offer, it becomes easier to contemplate about possible interventions for it is known, to a certain degree, how damage can be minimized. Comparing risks in the light of this knowledge is one way to come up with the proper tradeoff. In particular, the tradeoff is between those who are threatened by intervention A and those who are threatened by not making intervention A and what each threat includes given the theoretical and empirical background that we are aware of. Governments ought to have considered, for example, what risks are posed by the virus to the

4.2  Estimating Tradeoffs: Established Knowledge as Guiding Principle

77

population, what risks are posed by closing businesses and what risks are posed by leaving them open. The measures briefly mentioned in the introductory comment of this chapter, i.e., the YLL and the HRs are strongly linked both with disease-induced damage and economic-induced side effects. The YLL measure gives us information about premature deaths. Contrary to the absolute number of deaths from each specific cause which assigns the same weight to each death regardless of age, the YLL takes into account both the frequency of deaths and the age at which it occurs. So to estimate the burden of a disease and impact of public health policies, YLL should be the taken into account (WHO, 2022). To calculate YLL, we set an upper limit as a reference age and we sum the number of deaths at each age and multiply it by the number of years remaining to the reference age (Gardner & Sanborn, 1990). The calculation is done for deaths from each particular cause in each age-group and the results are summed. The average lifespan in western world is in the region of 78–85 years old (Human Development Report, 2020, 343) so say that we set 85 as the reference age. If someone dies when s/he is 80 years old the YLL are 5. If someone dies when is 50 years old, the YLL is 35. Now if a couple dies when they are both 30 years old due to a particular cause (say drug overuse), the YLL would be: 2 × 55 = 110. To apply the estimate to the covid crisis, recall that in the UK, the median age of deaths involving/ or due to Covid-19 is 83 years old and the mean age of deaths involving/or due to Covid-19 is 80.4 years old, while the average lifespan is 81 years old (ONS, 2021) So, by and large, there are hardly YLL due to Covid-19 in the UK. The side effects of lockdowns are likely to severely restrain the life span of young people and thus cause serious damage in the society in terms of YLL while Covid-19 does not do so. Another key measure to understand lockdown-induced damage is HRs. A typical characterization of HRs is that it is the ratio of death probabilities. Among two groups, the one has higher chances of dying in comparison to the other. For example, an HR of 3 means that the one group has a threefold increase in the chances of dying in comparison to the second group. So among those who face unemployment (a common consequence of lockdown policies) and those who do not, we try to estimate the extent to which the first group has considerably higher probability of dying prematurely. The side effects of school closures, the damage caused in the economy by the lockdowns, and the adverse effects of isolation are likely to result in many YLL and in higher HRs for young people. Undiagnosed and largely untreated diseases were also a side effect of mitigation measures leading to even more side effects. It is clear that policy makers ought to have considered the possibility of such outcomes prior to lockdown-implementation. Therefore, taking into account already established knowledge about respiratory viruses and about economic outcomes could have led them to figure out how YLL and HRs are associated with them which could have helped them to come up with a much more scientifically informative decision-making than the one they actually implemented. I mention so far mostly the side effects from the economic outcomes, but the damage caused by the mitigation measures is not confined to them; school closures and restricted socialization also have serious connotations which, again, count in YLL. Let us see how policy-making inflicted avoidable damage to the population by failing to consider the proper tradeoff.

78

4  Tradeoffs and Knock-On Effects

4.3 School Closures We have already seen the very low IFRs of Covid-19 in these age-groups (roughly 5–18), which is in the region of 0.00013%. This on its own could have been enough to argue against school closures; alas, that was not the case. As we saw, the three mistaken postulates upon which mitigation measures were based treat schoolchildren as people with relatively high risk of dying and hold that even if they have lower risk than the older ones they can transmit the virus to the elderly and thus lead to a remarkable increase in overall death rates. We have already seen that this was a mistaken thesis and that schoolchildren do not spread very much the virus for the simple reason that they barely ever develop symptoms which indicates that neither schoolchildren nor their close contacts benefited from school closures (Vlachos et al., 2021; Zsigmond et al., 2022). School closures therefore could make some sense if children died at high rates, which is not the case, or even if children in countries that closed schools died in much lower rates in comparison to countries that left their schools open. To the best of my knowledge, only Sweden did not close its schools (aged 1–16) at all throughout the pandemic, and so, based on pro-school-closures argument, Sweden’s Covid-19 death rates could have been higher than countries that closed schools and deaths and infections among schoolchildren could follow suit. We saw that the first part of this prediction does not hold when we discussed the failures of lockdowns and we have data which refute the second part too and show that schoolchildren in Sweden remained literally unaffected by the virus. In fact, Sweden faced zero covid-­ related deaths among schoolchildren (Carlson, 2020). In particular, researchers reported that among the 1.95 million children who were 1–16 years of age, 15 children had Covid-19, MIS-C, or both conditions and were admitted to an ICU, which is equal to 1 child in 130,000 and none of them died (Ludvigson, 2021). On top of that, no mask mandates were in place in Sweden which simultaneously casts further doubt not only on the need for school closures but moreover on universal masking policies (which included schools). These results emphatically debunk arguments in support of school closures for they show that children do not die of the virus and in combination with the findings already presented according to which they even rarely transmit the disease, it suggests that closing schools protects no one. However, one could argue that children with comorbidities could have benefited from school closures since they were at substantially higher risk of dying from the virus. It is not at all clear though that children with comorbidities were at much higher risk from the virus since, as we have seen, while it is accurate to say that some co-morbidities increase the risk of serious disease, age was the main determinant vis-à-vis covid outcomes. Indeed, in the UK, infections that occurred in immune-compromised children lead to no deaths (Chapel et al., 2021) or reported exceptionally rare deaths (Swann et al., 2020), which makes rather clear that possible underlying conditions in schoolchildren do not increase the risk of severe covid. But even if that was the case, and immune-compromised children were at higher risk than healthy children, then established knowledge on transmission

4.3  School Closures

79

dynamics as well as the lessons we could draw from the failure of lockdowns and similar mitigation measures, suggest that if these children were to be protected then, again, closing schools and shutting the country down could have offered little to no protection to them for the simple reason that, as we saw before, the immune-­ compromised children would take the same risks as the healthy ones and in that way they could have hardly been protected. Instead, to offer some protection, schools ought to have remained open, healthy children should have attended classes, and children with co-morbidities could perhaps self-isolate until some level of immunity was developed in the population. Thus, it is hard to see even an imaginary scenario whereby school closures are able to save lives. And the data on offer, drawn from the real world, as well the theoretical knowledge about transmission dynamics are quite clear and we can confidently draw the conclusion: Sars-Cov-2 should have never been an issue of concern with respect to schoolchildren either regarding healthy or immune-compromised ones and thus closing schools did not help them. We should therefore see whether it harmed them. The impact school closures had on children across the globe (with the exception of Sweden), counts in YLL. First, loss of educational attainment is linked to poorer lives and hence to lower lifespan. Virtual learning is not as effective as in-person learning (Jack et al., 2022; Goldhaber et al., 2022), especially for primary school children, and low levels of attendance make matters worse. According to estimates from the U.S., a total of 24.2 million children aged 5–11  years attended public schools that were closed during the 2020 pandemic, losing a median of 54 (interquartile range, 48–62.5) days of instruction. Missed instruction was associated with a mean loss of 0.31 years of final educational attainment for boys and 0.21 for girls. Summed across the population, 5.53 million of YLL may be associated with these school closures alone (Christakis et al., 2020). These are ground-breaking results that underscore the damage inflicted to children by school-closures. They are, moreover, scientifically sound; in contrast to the already discussed models on NPI which postulate undocumented claims (such as that large-scale quarantines work or that masks reduce transmission rates) and then go on and prove them, Christakis et al. postulate claims that are long-known to be quite valid and then estimate their impact with respect to the covid crisis. The correlation between education and health is positive,3 rather robust and cuts across both sexes (Kitagawa & Hauser, 1973; Grossman & Kaestner, 1997). The correlation between health and schooling in particular is one of the most well-­ established findings in social sciences (Conti et al., 2010), and some studies, known prior to the Covid-19 pandemic, estimated the gap in life expectancy between high school dropouts and those with at least a college degree to be 14.2 years for males and 10 years for females (Olshansky et al., 2012). Two RCT have vindicated these findings: the so-called Abecedarian Program (Campbell et al., 2014) and the Perry School program (Heckman et al., 2013) which showed that by providing intensive intervention to disadvantaged children health behaviors and health measures in

 It can also be equally described as a negative correlation between mortality and education.

3

80

4  Tradeoffs and Knock-On Effects

adulthood are significantly improved. Muney puts the premise “increased schooling leads to increased longevity” into test by conducting a quasi-natural experiment in US States that enforced compulsory schooling and by focusing on cohorts born between 1900 and 1925 found that children that attended school more, had lower likelihood of premature mortality in their adulthood (Muney, 2002). A number of studies that have been published over the years have consolidated this proposition, and some of them find that educational reforms often further enhance the positive correlation between the two variables (Davies et al., 2018).4 What explains this correlation remains unclear and scholars contemplate over several variables that are assumed to play a significant role. Among the possible explanations are that educated people are better informed decision-makers (Grossman, 1975), and some scientists attribute it to different attitudes among different individuals and assume that impatient people invest little in education in contrast to patient ones and thus the latter reap the fruits of their efforts later in life (Fuchs, 1982). Indeed, there seem to be a number of confluent factors (Muney, 2022), and not necessarily mutually exclusive ones, which perhaps include, apart from those already mentioned, socialization, which improves well-being (Berkman, 1995) and better career prospects, which typically includes better salaries (Olshansky et al., 2012). The mechanism that is at work beneath this correlation is not of paramount importance for our discussion and it suffices to stick to the fact that the proposition “educational attainment leads to better health outcomes” is accepted scientific knowledge which can be reliably used as a premise when modeling takes place and when projections about the impact of school closures are offered.5 And this is what Christakis et al. have done; as stated, they posited already established knowledge which was based on repeatedly confirmed evidence and estimated how this premise performs when applied to the Covid-19 pandemic. Empirical findings further enhance the hypothesis that school closures had devastating effects on schoolchildren without benefiting them by demonstrating the impact this policy had on their wellness. Children’s physical activity was severely reduced. A systematic meta-analysis revealed that schoolchildren suffered 20% decrease in physical activity (Neville et al., 2022), which in turn indicates that their health has already been damaged, for physical activity is a central ingredient to wellbeing and health, especially when it comes to children aged below 18 years old.

 However, it should be noted, that other findings suggest that compulsory educational reforms are of zero effect (Clark & Royer, 2013), though this does not alter the conclusion that schooling is strongly linked to a healthier life; it simply suggests that reforms of compulsory schooling do not necessarily improve health outcomes but it does not follow that the correlation between school attendance and longer lifespan no longer holds. Low school attendance can be a matter of fact even when schooling is compulsory—which it is in nearly all countries on earth—and thus while compulsory schooling is a good indicator to estimate school attendance, it does not rule out the possibility that the law is not enforced in many cases and that many children do not go to school. 5  A report published in mid-2022 by the OECD (2022) makes the case against school closures by providing evidence suggesting that closing schools was one of the biggest policy mistakes during the Covid-19 crisis. 4

4.4  Economic Devastation Part 1: Deep Recession

81

It is becoming therefore increasingly clear that when considering schools and schoolchildren the tradeoff was not between lives and education nor lives and socialization; long established scientific knowledge about the impact of education on health suggests that the tradeoff was between the lives that are risked by Sars-­ Cov-­2 and the lives that are risked by reduced educational attainment. It was very early on known that schoolchildren are extremely unlikely to die from the virus and that they very rarely even develop symptoms, and it has been known for decades that social scientists have offered findings that show that the toll of missing school is life-threatening. In Sweden where no lockdowns were in place and schools remained open, there were zero deaths from the virus among schoolchildren, though the evidence suggests that children in countries with closed schools are exposed to elevated risk of reduced life expectancy and YLL due to school closures. Thus, closing schools damaged children without benefiting them and this is not evidence-­ based policy making and is hardly compatible with a public interest explanation of school closures.

4.4 Economic Devastation Part 1: Deep Recession When making claims about the impact lockdowns and other mitigation measures had on economy (though lockdowns are the main source of such side effects), we need one more time to take into account the risks that are posed by an intervention, which in the case of lockdowns included primarily shutting down businesses and doling out money, in contradistinction to the risk that was posed by Sars-Cov-2. It appears that the economic outcomes of the lockdowns can be described as a three-­ staged process. The first two are a direct consequence of lockdowns. First, recession and increased unemployment rates and, secondly, huge public debts and increased inflation rates as a result of spending spree policies. Theory suggests that lockdown policy-making is the primary cause of both the former and also of the latter. The third one is a new round of recession in an attempt either to stop increased inflation rates or to eliminate deficits and repay the debts. By examining each one separately, the standard theoretical and empirical knowledge can be applied to show that lockdowns caused harm to huge segments of the society and in particular to those that faced little to no risk from the virus and to indicate one more time that lockdowns are not compatible with a public-interest theorizing of the pandemic policy-making. As lockdown policies were implemented during the year 2020 and much of the 2021,6 recession, mass unemployment, and reduced income became the norm. Across the world, the sharp rise in global poverty was unprecedented. To mention but a few results, a survey investigating the economic impact of the lockdown in

 In some places like in Netherlands and in Austria, lockdowns were enforced even in early 2022 and in China in mid-2022. 6

82

4  Tradeoffs and Knock-On Effects

Uganda found a decline in total household income of 60% (Mahmud & Riley, 2020). Reduced income in developing countries amounts to a shocking 67%, while the negative effects are experienced by households across the socioeconomic spectrum (Egger et al., 2021). On top of that, the World Food Program (WFP) of the United Nations warned that up to 270 million people were pushed to the brink of starvation during the pandemic (WFP, 2020). The empirical relationship between income and life expectancy is not new in economics and neither is the strong correlation between employment rates and lifespan. Samuel Preston’s famous curve illustrates that countries with higher average income have longer life expectancy than countries with lower average income and people within a country with higher income generally live longer than poorer ones (Preston, 1975). So economic suppression diminishes the income of many people and hence reduces the likelihood of living longer. The converse also holds. Economic growth is likely to increase the average lifespan by offering chances for higher income. There is moreover a clear correlation between employment rates and mortality. Data from the deep Swedish recession of 1992–1996 show that mass unemployment imposes mortality risk on a sizeable segment of the population. In particular, exposure to unemployment during recessions poses a higher mortality risk in post-recession periods. Post-recession, all-cause mortality was raised and gave Hazard Ratios (HRs) for men = 1.46 and women = 1.12. Unemployment is especially linked with alcohol-related health problems, ischemic episodes, and circulatory diseases. Excess deaths from circulatory diseases were recorded among men who experienced unemployment: from heart attacks HR = 1.32 and from ischemic episodes HR = 1.11. Young men experienced higher likelihood of death from stroke: HR  =  2.53. Alcohol-related mortality was significantly raised too, both among men and women. Among those in their 30s, i.e., people born from 1956 to 1965, these ratios were very high: for males HR = 4.44 and for females HR = 5.73. For men, but not for women, suicide and cancer mortality were also raised among those who had suffered unemployment (suicide HR = 1.43; cancer HR = 1.14). A significant result of this study is that mortality consequences are larger among the young, among the unmarried men and women and among men and women of low education or income (Vagero & Garcy, 2016).7 Researchers have offered estimates on the potential life-threatening effects of the 2020 recession. Using regression analysis, the Preston curve relationship between national income and childhood mortality is estimated at different levels of GDP fall. The analysis focuses on low income countries and on under-5 mortality. Under a conservative scenario (5% reduction on GDP per capita), the authors suggest that the total number of under-5 deaths increases to 19.5 million, or an additional 282,996 number of deaths (95% CI: 279,779–286,400). The results for each scenario at the country level suggest that for the scenarios of 10% and 15% GDP reductions, there is an estimated under-5 loss of life of 19.8 and 20.2 million, which  While the correlation between economic development and mortality may not be equally strong for all causes of death—for cardiovascular-related deaths are strongly procyclical but cancer-induced deaths may be unrelated to economic downturns or even countercyclical (Ruhm, 2013)—the association between higher income and longer life expectancy remains robust (Chetty et al., 2016). 7

4.5  Economic Devastation Part 2: Increased Spending and Inflationary Tolls

83

corresponds to an additional 585,802 (95% CI: 579,184–592,799) and 911,026 (95% CI: 900,804–921,825) lives lost, respectively. Moreover, we estimate that 49% of the total under-5 lives lost would occur in Sub-Saharan Africa, a pattern that is observed across the four scenarios, where the total number of lives lost in this region increased up to over 470,000 between a no downturn scenario and a 15% reduction in GDP per capita (Cardona et al., 2022). In a similar vein, economists project that lockdown policies are likely to result in excess deaths in the future. Covid-19-related unemployment was 2 and 5 times larger than the typical unemployment shock, depending on race/sex, which correlates with a 3.0% increase in mortality rate and a 0.5% drop in life expectancy over the next 15 years for the overall American population. Bianchi et al. predict that the shock will disproportionately affect African Americans and women, over a short horizon, while white men might suffer large consequences over longer horizons. These figures translate in a staggering 0.89 million additional deaths over the next 15 years (Bianchi et al., 2020). By bearing in mind that people who face unemployment are usually under 70 years old, we arrive at the following results: for those under 70 the Covid-19 infection fatality rate is = 0.05% while the mortality risk due to unemployment = 3% which is very high both in comparison to the low risk these age-groups face from the virus and on its own terms.8 A 3% mortality rate that would impact, say, on 100,000 people could lead to 3000 deaths while Covid-19’s death toll per 100,000 people for individuals below 70 years old amounts to about 50 deaths. These results imply that many YLL are to be expected and increased HRs are likely to be a matter of fact as a result of the lockdown-induced recession. Indeed, countries with higher average income are likely to suffer fewer excess deaths per million of the population in comparison to developing countries, but this does not alter the central conclusion: For people under 70, recession, fall in GDP and unemployment are all far more dangerous than Covid-19.

4.5 Economic Devastation Part 2: Increased Spending and Inflationary Tolls One of the most consequential facts in the aftermath of lockdowns is that deep recession co-occurred with increased public expenditures. Artificially low interest rates enabled governments in the EU, the USA, the UK, and the rest of the world to increase their spending exponentially. In the United States, the government spent  The spike in lockdown-induced unemployment rates lasted approximately from March 2020 till summer-fall 2021 (that’s the period of time in which repeated lockdowns were mostly implemented), while the unemployment mortality risk in Sweden lasted for four years. This suggests that individuals were exposed to the 3% unemployment mortality risk for a shorter period of time than people who suffered the Swedish recession, which could lead to fewer excess deaths in the period after the lockdown-induced recession than the ones recorded after the recession in Sweden. However, the case remains that they have been exposed to higher mortality risk than the risk posed by the virus. 8

84

4  Tradeoffs and Knock-On Effects

over 3.5 trillion additional dollars during the pandemic (USA Spending, 2021), while the national US debt climbed to $30 trillion (US Debt Clock, 2022). In the EU the lockdowns reversed the downward trend in government spending that was observed over the last ten years and increased it from 46.6% in 2019 to 53.4% in 2020 (Trading Economics, 2022a). In the UK, the public sector net debt (excluding public sector banks) was around 96% of GDP at the end of December 2021, which is the highest ratio since the 98.3% recorded in March 1963 (ONS, 2022a). In Africa, the government expenditure to GDP increased from 20.97% in 2019 to 22.25% in 2020 and 22.21% in 2021 (Statista, 2021). As these data demonstrate, a direct consequence of these spending spree policies is that they increase public debts. Increased public deficits are also a matter of fact. As the Eurostat mentions, in 2021, all member states (with the exceptions of Denmark and Luxemburg) reported deficits. The highest deficits were recorded in Malta (−8.0%), Greece (−7.4%), Latvia (−7.3%), Italy (−7.2%), Romania (−7.1%), Spain (−6.9%), Hungary (−6.8%), France (−6.5%), and Slovakia (−6.2%) (Eurostat, 2022). We can thus legitimately expect a long-lasting economic carnage across the globe in the years to come, as the citizens will be required to pay off these deficits and to reduce the increased debts. Lockdown advocates proposed that the economic outcomes would have been in the same ballpark even if the economies were not shut down, because people would self-isolate and thus the economic devastation is to be put down to the virus and not to the policy responses. This claim in fact extends and applies the public interest theory of lockdowns to the economic domain. Thus, it is typically a follow-up argument to the main pro-lockdown proposal that lockdowns saved millions of lives. The argument in unison was that lockdowns saved millions of lives and despite the fact that recession, unemployment and huge public debts became a matter of fact, this would have happened anyway but if it happens under lockdown, people are moreover saved from the virus.9 Since we have already documented the failures of lockdowns, it is enough to show that the part of the argument that is related to the economy is mistaken too. While it is true that people all over the world were under the sway of mass hysteria, as we have seen in Chap. 3, in no lockdown countries mass hysteria declined since people had the chance to estimate personal tradeoffs whereas in  lockdown states mass hysteria was magnified. Thus, in no lockdown states, like Sweden and South Dakota, the economic damage was minimized in comparison to pro lockdown countries. Unemployment rates in the United States verify that economies with no lockdowns perform far better than economies with

 Katz et al. (2021) try to make such a case in a publication in the BMJ. They recognize that there may be health-related lockdown-induced toll, though they suggest that an unmitigated epidemic would pose far greater harm. Their argument hinges on the three mistaken pro-lockdown premises discussed in Chap. 2, and as for the side effects of the mitigation measures, they claim that it is difficult to evaluate which side effects are to be attributed to lockdowns. However, as I show in this chapter, by having established knowledge as guiding principle and by paying attention to the data that emerged, we can be confident about the damage inflicted by the lockdowns and make causal claims thereof. 9

4.5  Economic Devastation Part 2: Increased Spending and Inflationary Tolls

85

lockdowns. In December 2021, South Dakota had among the lowest unemployment rates in the United States (2.6%), and it is, to the best of my knowledge, the only state in the country that did not shut down businesses. On the contrary, New York, where very harsh lockdowns were in place, had, during the same period, more than double the unemployment rates of South Dakota (6.6%). The situation was similar in California, which had the highest unemployment rates in the country (6.9%) and also implemented strict repeated lockdowns (BLS, 2021). These gaps were even higher in spring 2020: unemployment rates in South Dakota were 8.8%, in New York 16.2% and in California 16% (BLS, 2022). The Eurostat piece cited above further strengthens the claim that lockdown is the cause for the worsening economic conditions, not the fact that a novel pathogen circulated. By the end of 2021, Sweden was among the countries with the lowest ratios of government debt to GDP in the EU. In particular, Sweden’s rate was 36.7% while 14 member states had government debt ratios higher than 60% of GDP, with the highest registered in Greece (193.3%), Italy (150.8%), Portugal (127.4%), Spain (118.4%), France (112.9%), Belgium (108.2%), and Cyprus (103.6%). On top of that, Sweden had the lowest recession in the EU (2.9%) (European Commission, 2022), while the average estimate in the EU was 6.3%, in Eurozone 6.8% (European Commission, 2021, 17) and in the UK 9.9% (ONS, 2022c). Thus, mass hysteria may be linked with some hesitancy in consumption for a period of time and it perhaps delays some investments, but it does not cause deep recession, mass unemployment and rapid increase in public expenditures—lockdowns do so. It has been argued that high levels of economic development would ensue the lockdown-induced recession and will suffice to eliminate the deficits and to reduce the debts. This claim, which at first sight seems to be based on Keynes’s theory, is undercut by the second major issue that appears with a policy-making that increases public expenditures, namely that it exposes individuals to inflationary tolls. Thus, Keynes’s theory does not find application because Keynes did not suggest that governments should borrow and print money and hand it out to closed businesses and furlough or unemployed individuals; instead, he said that public spending should take place when private investments are inefficient at enhancing consumer spending and thus stabilize aggregate demand (Keynes, 1972, 1973).10 In a closed economy, this is impossible to occur because investments (either private or public) are non-­ existent, and it is rather unlikely to boost consumption under such circumstances. Instead, deep recession emerges and many people stay at home and hold money that came out of thin air. It is hard to see a reasonable Keynesian pathway here. On the contrary, this is a prescription for inflation and Keynes was fully aware that this can happen lest public spending gets out of control and strongly recommended that  This is in contrast with the Austrian theory of economics which suggests that individuals are the ones who have the knowledge and the motivations to bring the economy to equilibrium. I do not attempt a comparison, let alone an evaluation, of the two theories here. I simply stress that Keynesian economics are irrelevant to the Covid-19 pandemic and thus those who advocate a public-interest approach to the covid-crisis do not make a well-thought through assessment if they rely on Keynes’s writings as a confirmation of their theory. 10

86

4  Tradeoffs and Knock-On Effects

spending should not be left unadjusted for he was mindful that inflation does irreparable damage to a society (Humphrey, 1981). Austrian economists could object that there is no need to misapply Keynes’s theory to have inflation, for Keynes’s approach per se paves the way to inflation. But to put Keynes’s theory into test, then governments need to do at least approximately what Keynes endorsed. However, by shutting everything down and by increasing relentlessly their expenditures, it turns out that what governments did during the covid crisis had no basis on any school of thought in economics, either this is Keynesian, Austrian, or other. Therefore, the inflationary trends that came in immediately after lockdowns were lifted (from May 2021 onwards, in most parts of the world), is what it was reasonable to expect in view of some well-known economic premises of the quantity theory of money, namely that when the stock of money increases prices need to follow suit to keep the economy in equilibrium. The post-lockdown inflation was primarily demand-driven and was not a result of supply constraints. Governments and lockdowners that cherished the idea of a “transitory” inflation, relied on the assumption that inflation was a result of problems in the global supply chain (which they took them to be unrelated to lockdowns), and once these problems are resolved, inflation will no longer exist. However, inflation persisted for more than a year, increased month after month, and turned out to being all but “transitory.” This brings us to an important distinction that needs to be made—that between relative price changes and general price changes—and which leads to the conclusion that inflation is nearly always the result of increasing the availability of money in the society.11 Relative price changes and general prices changes are explained by different mechanisms and while economists have distinguished between them, this is a point that is constantly missed by policymakers. Greenwood and Hanke draw insights from the quantity theory of money and examine three case studies which show that relative price changes happen all the time and are related to supply issues in the real economy while the overall price increase is related to the amount of money that is available. Relative price changes are thus irrespective of the overall inflation rate (Greenwood & Hanke, 2022). Therefore, had the inflation been supply-driven, we would have observed prices to rise in some goods and services relative to others and in particular to those goods that were directly affected by the supply chain drawbacks because the same amount of money would compete to buy fewer amounts of the goods in question; instead, what happened gradually since spring-summer 2021, was that the prices of all goods12 started to spike, exactly because new money was

 Milton Friedman famously said that inflation is always and everywhere a monetary phenomenon (Friedman, 1956). 12  Services such as hospitality are lagging behind, reversing the trend that lasted until 2019. The reason is that increased money supply shifted consumer demand into goods and lesser into services which consolidates the distinction between relative price changes and the overall inflation rate. This trend is similar in the USA and the Eurozone while in the UK prices in services increase alongside prices in goods. This is due to the blurred boundaries between goods and services in the UK, which is quintessential to the UK’s economy, for example, many goods are sold as “bundled” services, such as “services” in the supermarket (Greenwood & Hanke, 2022, 44). 11

4.5  Economic Devastation Part 2: Increased Spending and Inflationary Tolls

87

entered and competed to buy pretty much the same amount of goods. The data are in: In June 2021, inflation rates in the USA were at 5.4% and this trend continued unabated for more than a year. In January 2022, the rate was 7.5%, and in May 2022, it was 8.5% (Trading Economics, 2022b). Likewise, in the EU, in June 2021, the inflation rate was 2.2%, in January 2022, it climbed to 5.6%, and in May 2022, it was 8.8% (Trading Economics, 2022c). In the UK, the inflation rate in April 2021 was 1.6% whereas in April 2022 it was 7.8% (ONS, 2022b). The evidence is compatible with the quantity theory of money and rejects the “transitory” approach to post-lockdown inflation. Persisting inflation that increases on a monthly basis (which is what it is typically meant when people use this terminology), halts economic growth and so the argument that debts and deficits can be reduced in a (pseudo)Keynesian manner is not a plausible scenario for the post-lockdown economy. Robert Barro offered cross-­ country analysis of 100 countries, in which he examined the relationship between inflation and economic growth. The findings indicate that an increase in the average inflation rate by 10% per year is estimated to lower the growth rate of real per capita GDP on impact by 0.3–0.4 percentage points per year (Barro, 1996). Since the best part of the global economy was in deep recession approximately from March 2020 till May 2021, the inflationary rates that ensued suggest that the expected economic growth was impeded. Moreover, the lockdown-recession was great and the economic bounce back did not fully restore the rates to pre-pandemic levels. By adding the effects of inflation to this, we can conclude that while by lifting lockdowns recession was mostly reduced or even eliminated, the great economic growth that could cure the harms of recession was never documented and inflation could have played a part in this. Data from the EU show that during the second quarter of 2020 the economic downturn was 14.6% while the second quarter of 2021 the outcome was 14.4%. This leaves a lag of 0.2%. By the end of 2021 (fourth quarter), the economic growth was 4.6% and in the respective quarter of 2020 the downturn was 4.4% (Statista, 2022) which implies that while there was no more recession at that period of time, the reported economic growth was much lower than expected, i.e., 0.2%, and it appears that it cannot be argued convincingly that such rates are enough to eliminate the lockdown-induced debts and deficits. Thus, either in the near or in the distant future fiscal rules will likely take place in countries that implemented lockdowns in an attempt to eliminate deficits and reduce debts which in turn paves the way to another deep recession and to its life-threatening connotations. In case inflation persists well beyond the foreseeable future, increased interest rates can be ignited by central banks so that money supply will be reduced and inflation rates will do so accordingly. Again, this will almost certainly result in deep recession for it is akin to enforcing austerity measures. It has been proposed though, that inflation is not as bad as we may think. The well-known Phillips curve suggests that some level of inflation is optimal for it is linked with reduced unemployment rates. By applying this to the covid crisis therefore, one may argue that even if inflation is not transitory and it is persistent (whichturned out to be the case), then we could expect reduced unemployment rates in the long run. Milton Friedman has opposed the underlying premises of the Phillips

88

4  Tradeoffs and Knock-On Effects

curve according to which governments could reduce unemployment rates by artificially increasing the inflation rates. According to Friedman inflation could be linked to reduced unemployment only insofar as the inflation rates remain stable (say at 2%). If inflation rates increase then unemployment rates will also increase (Friedman, 1968, 8–10). The increase in unemployment rates in the USA from the mid-1960s to 1970s provides an empirical refutation of the Phillips curve prediction and a vindication of Friedman’s criticism. In 1964, the inflation was 1% and the unemployment rates were 5%. In mid-1970s, inflation was 12% and unemployment above 7%, while in mid-1980s, the inflation was 14.5% and unemployment at about 7.5% (Federal History Reserve, 2013). Thus, we have no compelling theoretical and/or empirical reasons to expect that the inflation rates that emerged after the lockdowns would reduce unemployment. On the contrary, we have good reasons to assume that increasing inflation rates will increase unemployment rates too. Apart from impeding economic growth and tending to increase unemployment, inflation has moreover some not so well-known side effects, as for example that it is linked to increased homicide rates. The mechanism that establishes the link is not clear and it is assumed that acquisitive crime is a central ingredient of it; prior research suggested that inflation boosts acquisitive crime (Land & Felson, 1976; Rosenfeld & Levin, 2016) which in turn leads to increased homicide rates. Researchers examined homicide rates and inflation rates in several US cities and report striking correlation between increases in homicides alongside with increased inflation rates and conversely, reduced inflation rates with reduced homicides (Rosenfeld & Vogel, 2021). Persisting inflation therefore is linked to several life-­ threatening outcomes and offers no clear benefit to the majority of the population.

4.6 Isolation Social scientists have long demonstrated that social relationships are a key factor to mortality and have provided evidence indicating that those with stronger social relationships have a 50% increased likelihood of survival than those who live isolated (Holt-Lunstad et al., 2010). Apart from shutting everything down, lockdown policies involve work-from-home mandates for almost every employee. Working from home and studying from home while curfews are still in place-which means that there is no room for socialization- leads many people to isolation. Thus, almost all people, either those who lost their job or those who simply worked or studied from home (including schoolchildren and students), were forced to isolation. Data from Switzerland indicate that the average person would suffer 0.205 YLL due to psychosocial consequence of Covid-19 mitigation measures, i.e., both unemployment per se plus the lack of social relationships. However, this loss would be entirely borne by 2.1% of the population, who will suffer an average 9.79 YLL. According to the authors, these are moderate estimates (Moser et al., 2020). These results correlate with further empirical research that has been made. Empirical investigation shows increased social anxiety among young people. Researchers examined how

4.7  Unreasonably (?) High Excess Deaths

89

mandated low social exposure influenced socially anxious university students, and compared their anxiety to that of socially anxious students in preceding academic years with no social distancing. The results showed that whereas social anxiety decreased in socially anxious students from the fall to the spring semester in the years preceding the pandemic, during the 2019–2020 academic year (which involves lockdowns followed by social distancing measures at the end of the fall semester), social anxiety levels remained high and unchanged (Arad et al., 2021). The CDC discusses the results of a study released by the National Academies of Sciences, Engineering, and Medicine in which the impact of isolation on individuals’ well-being is well documented. It verifies that it increases anxiety, that it is linked to suicidal ideation, and that it furthermore increases the risk of heart disease. Most strikingly, loneliness among heart failure patients was associated with a nearly 4 times increased risk of death, 68% increased risk of hospitalization, and 57% increased risk of emergency department visits (Federal History Reserve, 2013).

4.7 Unreasonably (?) High Excess Deaths Some well-established scientific postulates from economics and social sciences indicate that by disrupting economic and social life at such lengths as it happened with the lockdowns, this exposes people to serious death threats. While, as we have seen, the most severe consequences are expected in the years to come through increased HRs and YLL, there are some findings which suggest that the impact of lockdowns was so great and multi-layered that the toll can be even worse than the standard theoretical and empirical knowledge could predict. And while this may seem surprising at first sight, in fact it may not be so. Sweeping interventions of this sort on a global scale and for such a long period of time have never occurred in the past and so a precise estimate for the damage that inflicted is hard to make for there are confluent factors that appear to exacerbate the effect. One such factor is the delayed diagnoses of serious diseases which took place in several countries under lockdown. In the New England, a significant decrease in cancer screenings was reported during the first lockdown, i.e., from March 2020 through June 2020, while from June 2020 through September 2020, there was a rise in the number of screening tests (Bakouny et al., 2021). A comparison of patient encounter data between April 2019 and April 2020 show a remarkable decline in both existing and new incidence neoplasms (malignant, benign, in situ, and unspecified behavior) as well as substantial decreases in breast cancer (−89.2%) and colorectal cancer (−84.5%) screenings (London et  al., 2020). In India, between March 1 and May 31, 2020, the number of new patients registered decreased from 112,270 to 51,760 (54% reduction), patients who had follow-up visits decreased from 634,745 to 340,984 (46% reduction), hospital admissions decreased from 88,801 to 56,885 (36% reduction), outpatient chemotherapy decreased from 173,634 to 109,107 (37% reduction), the number of major surgeries decreased from 17,120 to 8677 (49% reduction), minor surgeries from 18,004 to 8630 (52% reduction),

90

4  Tradeoffs and Knock-On Effects

patients accessing radiotherapy from 51,142 to 39,365 (23% reduction), pathological diagnostic tests from 398,373 to 246,616 (38% reduction), number of radiological diagnostic tests from 93,449 to 53,560 (43% reduction), and palliative care referrals from 19,474 to 13,890 (29% reduction). These reductions were even more marked between April and May, 2020 (Ranganathan et al., 2021). Despite the fact that it was known in theory that delayed diagnoses of major-killer diseases damages public health, this was mainly related to the overall economic status of a country. For example, while that was to be expected to happen in the developing world, it was nearly unimaginable to happen en masse in places like the US. And if this happened in the West during lockdowns, it is fair to assume that it also happened in the developing world and perhaps led to even worse outcomes. One could reply that the governments never prohibited people from attending hospitals and so it is their own fault that they did not pay a visit to a hospital. It is true that, at least as far as I am concerned, people were not forbidden from visiting hospitals during lockdowns, though lockdowns are still primarily responsible for this outcome. We have seen that people were panicked and behaved irrationally during the pandemic and that the panic was magnified due to lockdowns. This made them being afraid of covid much more than of cancer or of heart-attacks. Avoiding covid at all costs was the main mantra among the public and the politicians in lockdown countries and the costs in part included missing treatments or delaying diagnoses. On the contrary, if no lockdowns were in place, people would have gradually realized that they could keep up with their daily schedule without much trouble, and, among other things, visits to hospitals would have not seen such a sharp decrease. Deaths of despair (drug overuse, alcohol, etc.) (Mulligan, 2020) is another factor that whilst it is predicted by the increased HRs that are expected due to recession, reports suggest that alcohol and drug overuse led to deaths earlier than anticipated. Data from countries that locked down still pouring in and show a 40% increase in overdose-associated cardiac arrests in the USA (Friedman et  al., 2021) while in Scotland the number of alcohol-specific deaths has increased by 17% to 1190 in 2020, up from 1020 in 2019 (National Records of Scotland, 2021). The literature appears to be inconclusive as for the impact of lockdowns on suicide rates during the recession, i.e., during 2020. For example, reduced suicides by 5.4% were recorded in the USA in 2020 (Ahmad & Anderson, 2021) while substance use and suicidal ideation was increased over the same period (CDC, 2020). However, as already noted, we already see excess deaths due to drug and alcohol overuse and due to cardiac arrest. It is likely that these figures, as well as suicide rates, will have an upward trend in the years to come in the aftermath of recession. When these results are considered alongside the serious economic downturn, this seems to pave the way for below par public health outcomes. Therefore, we should not be taken aback from the fact that excess mortality (i.e., covid and non covid deaths) is higher in places with lockdowns than in places without lockdowns. In Chap. 2, I described a possible mechanism that can perhaps lead to increased covid deaths under lockdown, but excess deaths cannot and are not attributed solely to the virus. The side effects of the mitigation measures appear to make the difference. The Center of Evidence Based Medicine which belongs to the Oxford University

References

91

published an index on excess mortality in 2020 which shows that most countries with lockdowns had more excess deaths than countries with no lockdowns. The excess mortality in Sweden was 1.5% and in South Korea was −2.9% (which means that there were no excess deaths), while the vast majority of lockdown countries reported excess deaths well above these figures and many of them go even higher than 10% (Parildar et al., 2021). A state-to-state comparison in the USA offers similar results. The overall mortality increases in the immediate weeks following the implementation of NPI (Agrawal et al., 2022). These data are largely compatible with the theoretical and empirical knowledge that was available prior to the pandemic and stress one more time that the tradeoff was lives against lives instead of lives against other valuable things, as a public interest explanation would propose and as many cost-benefit analyses posited.

References Agrawal, V., Sood, N., & Waley, C.  M. (2022). The impact of Covid-19 pandemic and policy responses on excess mortality (Working paper no. 28930). National Bureau of Economic Research. https://doi.org/10.3386/w28930 Ahmad, F. B., & Anderson, R. N. (2021). The leading causes of death in the US for 2020. JAMA, 325(18), 1829–1830. https://doi.org/10.1001/jama.2021.5469 Allen, D. (2021). Covid lockdown cost/benefits: A critical assessment of the literature. International Journal of Economics and Business. https://doi.org/10.1080/13571516.2021.1976051 Anderson, H., & Treich, N. (2011). The value of a statistical life. In A. de Palma, R.  Lindsey, E. Quinet, & R. Vickerman (Eds.), Handbook in transport economics (pp. 396–424). Edward Elgar Publishing. Arad, G., Leshem, D. S., & Bar-Harim, Y. (2021). Social distancing during a COVID-19 lockdown contributes to the maintenance of social anxiety: A natural experiment. Cognitive Therapy and Research, 45, 708–714. https://doi.org/10.1007/s10608-­021-­10231-­7 Bakouny, Z., Paciotti, M., et  al. (2021). Cancer screening tests and cancer diagnoses during the Covid-19 pandemic. JAMA Oncology, 7(3), 458–460. https://doi.org/10.1001/ jamaoncol.2020.7600 Barro, R. J. (1996). Determinants of economic growth: A cross-country empirical study (Working paper no 5698). National Bureau of Economic Research. Berkman, L. F. (1995). The role of social relations in health promotion. Psychosomatic Medicine; Journal of Bio-behavioral Medicine, 57(3), 245–254. Bianchi, F., Bianchi, G., & Song, D. (2020). The long-term impact of the Covid-19 unemployment shock on life expectancy and mortality rates (NBER working paper no. 28304). National Bureau of Economic Research. BLS. (2021). Unemployment rates for states. https://www.bls.gov/web/laus/laumstrk.htm. Accessed Dec 2021. BLS. (2022). Local area unemployment statistics. www.bls.gov/web/laus/lauhsthl.htm. Accessed 22 Jan 2022. Broome, J. (2008). Weighing lives. Oxford University Press. Campbell F, Conti G., Heckman J.J., Moon S., et.al. (2014). Early childhood investments substantially boost adult health. Science. 343 (6178): 1478–1485. Cardona, M., Millward, J., Gemmill, A., Jison, K., et  al. (2022). Estimated impact of the 2020 economic downturn on under-5 mortality for 129 countries. PLoS One. https://doi.org/10.1371/ journal.pone.0263245

92

4  Tradeoffs and Knock-On Effects

Carlson, J. (2020). Covid-19 in schoolchildren: A comparison between Finland and Sweden. Public Health Agency of Sweden. www.folkhalsomyndigheten.se/publicerat-­material. Accessed 2 Apr 2020. CDC. (2020). Mental health, substance use, and suicidal ideation during the COVID-19 pandemic—United States, https://doi.org/10.15585/mmwr.mm6932a1externalicon. Accessed 13 Aug 2020. Chapel, H., Patel, R., Driessens, C., Tarr, A. W., et al. (2021). Immunocompromised children and young people are at no increased risk of severe Covid-19. Journal of Infection, 84(1), 35–39. https://doi.org/10.1016/j.jinf.2021.11.005 Chetty, R., Stepner, M., Abraham, S., Lin, S., Scuderi, B., Turner, N., Bergeron, A., & Cutler, D. (2016). The association between income and life expectancy in the United States, 2001–2014. JAMA, 315(16), 1750–1766. Christakis, D., Van Cleve, W., & Zimmerman, F. J. (2020). Estimation of US children’s educational attainment and years of life lost associated with primary school closures during the coronavirus disease 2019 pandemic. JAMA Network Open, 3(11), e2028786. Clark, D., & Royer, H. (2013). The effect of education on adult mortality and health: Evidence from Britain. American Economic Review, 103(6), 2087–2120. Conti, G., Heckman, J., & Urzua, S. (2010). The education-health gradient. American Economic Review, 100(2), 234–238. Davies, N. M., Dickson, M., Smith, G. D., Van Den Berg, G. J., & Windmeijer, F. (2018). The causal effects of education on health outcomes in the UK biobank. Nature Human Behavior, 2(2), 117–125. Drummond, M. F., O’Brien, B., Stoddart, G. L., & Torrance, G. W. (1997). Methods for the economic evaluation of health care programmes. Oxford Medical Publications. Egger, D., Miguel, E., Warren, S. S., Shenoy, A., et al. (2021). Falling living standards during the COVID-19 crisis: Quantitative evidence from nine developing countries. Science Advances, 7(6), eabe0997. European Commission. (2021, February). European economic forecast (Institutional paper 144). European Commission. (2022). Economic forecast for Sweden. https://ec.europa.eu/info/business-­ economy-­euro/economic-­performance-­and-­forecasts/economic-­performance-­country/sweden/ economic-­forecast-­sweden_en. Accessed 16 May 2022. Eurostat. (2022). Provision of deficit and debt data for 2021 – First notification. https://ec.europa. eu/eurostat/documents/2995521/14497748/2-­22042022-­AP-­EN.pdf/332c1346-­5604-­bad3-­ adf1-­1fde9c4574f2. Accessed 22 April 2022. Federal History Reserve. (2013). The great inflation. https://www.federalreservehistory.org/ essays/great-­inflation. Accessed 22 Nov 2013. Friedman, M. (1956). The quantity theory of money-a restatement. In M. Friedman (Ed.), Studies in the quantity theory of money (pp. 1–21). University of Chicago Press. Friedman, M. (1968). The role of monetary policy. The American Economic Review, LVIII(3), 1–17. Friedman, J., Mann, N. C., Hansen, H., et al. (2021). Racial/ethnic, social, and geographic trends in overdose-associated cardiac arrests observed by US emergency medical services during the COVID-19 pandemic. JAMA Psychiatry, 78(8), 886–895. https://doi.org/10.1001/ jamapsychiatry.2021.0967 Fuchs, V. R. (1982). Time preference and health: An exploratory study. In V. Fuchs (Ed.), Economic aspects of health. The University of Chicago Press. Gardner, J. W., & Sanborn, J. S. (1990). Years of potential life lost (YPLL) – What does it measure? Epidemiology, 1(4), 322–329. Goldhaber, D., Kane, T. J., McEachin, A., Morton, E., Patterson, T., & Staiger, D. O. (2022). The consequences of remote and hybrid instruction during the pandemic (Working paper 30010). National Bureau of Economic Research. Greenwood, J., & Hanke, S. (2022). On monetary growth and inflation in leading economies, 2021–2022: Relative prices and the overall price level. Journal of Applied Corporate Finance, 33(4), 39–51.

References

93

Grossman, M. (1975). The correlation between health and schooling. In N.  E. Terleckyj (Ed.), Household production and consumption (Studies in income and wealth, Vol. 40, Conference on Research in Income and Wealth). Columbia University Press for the National Bureau of Economic Research. Grossman, M., & Kaestner, R. (1997). Effects of education on health. In J. R. Berhman & N. Stacey (Eds.), The social benefits of education. Stacey University of Michigan Press. Hammitt, J. K. (2020). Valuing mortality risk in the time of COVID-19. SSRN (Working paper. 3615314.) Heckman, J., Pinto, R., & Savelyev, P. (2013). Understanding the mechanisms through which an influential early childhood program boosted adult outcomes. American Economic Review, 103(6), 2052–2086. Holt-Lunstad, J., Smith, T.  B., & Layton, B.  J. (2010). Social relationships and mortality risk: A meta-analytic review. Plos Medicine, 7(7), e1000316. https://doi.org/10.1371/journal. pmed.1000316 Human Development Report. (2020). The next frontier. United Nations Development Program. http://hdr.undp.org/sites/default/files/hdr2020.pdf. Accessed 5 Dec 2020. Humphrey, T. (1981, January–February). Keynes on inflation. Economic Review, 3–13. Jack, R., Halloran, C., & Okun, J. (2022). Pandemic schooling mode and student test scores: Evidence from U.S. school districts (NBER working paper no. 29497). Katz, G. M., Bhatt, S., Ratmann, O., et al. (2021). Is the cure really worse than the disease? The health impacts of lockdowns during COVID-19. BMJ Global Health, 6, e006653. https://doi. org/10.1136/bmjgh-­2021-­006653 Keynes, J.  M. (1972). Essays in persuasion (The collected writings of J.M.  Keynes) (Vol. IX). Macmillan. Keynes, J. M. (1973). The general theory and after part IH. Defence and development (The collected writings of J.M. Keynes) (Vol. XIV). Macmillan. Kitagawa, E. M., & Hauser, P. M. (1973). Differential mortality in the United States: A study in socioeconomic epidemiology. Harvard University Press. Knight, F. (1921). Risk, uncertainty and cost. The Riverside Press. Land, K. C., & Felson, M. (1976). A general framework for building dynamic macro social indicator models: Including an analysis of changes in crime rates and police expenditures. American Journal of Sociology, 82(3), 565–604. London, J.  W., et  al. (2020). Effects of the COVID-19 pandemic on cancer-related patient encounters. JCO Clinical Cancer Informatics, 4, 657–665, https://ascopubs.org/doi/10.1200/ CCI.20.00068 Ludvigson, J. F. (2021). Open schools, Covid-19, and child and teacher morbidity in Sweden. New England Journal of Medicine, 384, 669–671. Mahmud, M., & Riley, E. (2020). Household response to an extreme shock: Evidence on the immediate impact of the Covid-19 lockdown on economic outcomes and well-being in rural Uganda. World Development, 140, 105318. Miles, D. K., Stedman, M., & Heald, A. H. (2020). “Stay at home, protect the National Health Service, save lives”: A cost benefit analysis of the lockdown in the United Kingdom. The International Journal of Clinical Practice. https://doi.org/10.1111/ijcp.13674 Moser, D., Glaus, J., Frangou, S., & Schechter, D. (2020). Years of life lost due to the psychosocial consequences of COVID-19 mitigation strategies based on Swiss data. European Psychiatry, 63(1), E58. https://doi.org/10.1192/j.eurpsy.2020.56 Mulligan, M. (2020). Deaths of despair and the incidence of excess mortality in 2020 (Working paper. 28303). National Bureau of Economic Research. Muney, A. L. (2002). The relationship between education and adult mortality in the United States (NBER working paper 8986). National Bureau of Economic Research. Muney, A.  L. (2022). Education and income gradients in longevity: The role of policy (NBER working paper 29694). National Bureau of Economic Research.

94

4  Tradeoffs and Knock-On Effects

National Records of Scotland. (2021). Alcohol-specific deaths in Scotland increase. https://www. nrscotland.gov.uk/node/3596. Accessed 17 Aug 2021. Neville, R. D., Lakes, K. D., Hopkins, W. G., & Tarantino, G. (2022). Global changes in child and adolescent physical activity during the COVID-19 pandemic. JAMA Network. https://doi. org/10.1001/jamapediatrics.2022.2313 OECD. (2022). Education recovery after COVID-19: Better, stronger & collaborative. https:// oecdedutoday.com/education-­recovery-­after-­covid/. Accessed 1 July 2022. Olshansky, S. J., Antonucci, T., Berkman, L., Binstock, R. H., et al. (2012). Differences in life expectancy due to race and educational differences are widening, and many may not catch up. Health Affairs, 31(8), 1803–1813. ONS. (2021). Average age of those who had died with Covid-19. https://www.ons.gov.uk/aboutus/ transparencyandgovernance/freedomofinformationfoi/averageageofthosewhohaddiedwithcovid19. Accessed 11 Jan 2021. ONS. (2022a). Public sector finances. https://www.ons.gov.uk/economy/governmentpublicsectorandtaxes/publicsectorfinance/bulletins/publicsectorfinances/december2021. Accessed 22 Feb 2022. ONS. (2022b). Consumer price inflation, UK: April 2022. https://www.ons.gov.uk/economy/inflationandpriceindices/bulletins/consumerpriceinflation/april2022. Accessed 22 June 2022. ONS. (2022c). Understanding the UK economy. https://www.ons.gov.uk/economy/nationalaccounts/articles/dashboardunderstandingtheukeconomy/2017-­02-­22. Accessed 24 June 2022. Parildar, U., Perara, R., & Oke, J. (2021). Excess mortality across countries in 2020. The Centre for Evidence-Based Medicine. https://www.cebm.net/covid-­19/excess-­mortality-­across-­ countries-­in-­2020/?fbclid=IwAR13Z-­7frMGM8BbE9yRHtaC0xxd8L2HQKAD63ULlDrP-­ Q1e17OLMscEDCnk. Accessed 3 Mar 2021. Preston, S. (1975). The changing relation between mortality and level of economic development. Population Studies, 29(2), 231–248. Prieto, L., & Sacristan, J. (2003). Problems and solutions in calculating quality-adjusted life years (QALYs). Health and Quality of Life Outcomes, 1(80). https://doi.org/10.1186/1477-­7525-­1-­80 Ranganathan, P., Sengar, M., Chinnaswamy, G., Agrawal, G., et al. (2021). Impact of COVID-19 on cancer care in India: A cohort study. Lancet Oncology, 22, 970–976. Robinson, L., Sullivan, R., & Shogren, J. (2020). Do the benefits of COVID-19 policies exceed the costs? Exploring uncertainties in the age–VSL relationship. Risk Analysis: An International Journal. https://doi.org/10.1111/risa.13561 Rosenfeld, R., & Levin, A. (2016). Acquisitive crime and inflation in the United States: 1960–2012. Journal of Quantitative Criminology, 32(3), 427–447. Rosenfeld, R., & Vogel, M. (2021). Homicide, acquisitive crime, and inflation: A City-level longitudinal analysis. Crime and Delinquency. https://doi.org/10.1177/00111287211039994 Ruhm, C. (2013). Recessions, healthy no more? (NBER working paper no. 19287). National Bureau of Economic Research. Schelling, T. C. (1968). The life you save may be your own. In S. B. Chase Jr. (Ed.), Problems in public expenditure analysis (pp. 127–162). Brookings Institution. Statista. (2021). Sub-Saharan Africa: Ratio of government expenditure to gross domestic product (GDP) from 2016 to 2026. https://www.statista.com/statistics/805575/ratio-­of-­government-­ expenditure-­to-­gross-­domestic-­product-­gdp-­in-­sub-­saharan-­africa/. Accessed 24 Nov 2021. Statista. (2022). Growth of the real gross domestic product (GDP) in the Euro area from 4th quarter 2018 to 4th quarter 2021. https://www.statista.com/statistics/226122/gdp-­growth-­in-­ the-­eu-­and-­the-­euro-­area-­compared-­to-­same-­quarter-­previous-­year/. Accessed 22 Mar 2022. Swann, O. V., Holden, K. A., Turtle, L., Pollock, L., et al. (2020). Clinical characteristics of children and young people admitted to hospital with Covid-19 in United Kingdom: Prospective multicentre observational cohort study. The British Medical Journal, 370, m3249. https://doi. org/10.1136/bmj.m3249 Thunström, L., Newbold, S., Finnoff, D., Ashworth, M., & Shogren, J. F. (2020). The benefits and costs of flattening the curve for COVID-19. Journal of Benefit-Cost Analysis, FirstView.

References

95

Trading Economics. (2022a). European Union Government spending to GDP. https://tradingeconomics.com/european-­union/government-­spending-­to-­gdp Trading Economics. (2022b). United States, inflation rate. https://tradingeconomics.com/ united-­states/inflation-­cpi Trading Economics. (2022c). European Union, inflation rate. https://tradingeconomics.com/ european-­union/inflation-­rate US Debt Clock. (2022). https://www.usdebtclock.org/ USA Spending. (2021). https://www.usaspending.gov/. Accessed 28 Feb 2021. Vagero, D., & Garcy, A. (2016). Does unemployment cause long-term mortality? Selection and causation after the 1992–96 deep Swedish recession. European Journal of Public Health, 26(5), 778–783. Vlachos, J., Hertegard, E., & Svaleryd, H. B. (2021). The effects of school closures on SARS-­ CoV-­2 among parents and teachers. PNAS. https://doi.org/10.1073/pnas.2020834118. Accessed 11 Feb 2021. WFP. (2020). WFP chief warns of grave dangers of economic impact of coronavirus as millions are pushed further into hunger. https://www.wfp.org/news/wfp-­chief-­warns-­grave-­dangers-­ economic-­impact-­coronavirus-­millions-­are-­pushed-­further-­hunger. Accessed 17 Sept 2020. Whitehead, S. J., & Ali, S. (2010). Health outcomes in economic evaluation: The QALY and utilities. British Medical Bulletin, 96(1), 5–21. WHO. (2022). Years of life lost (per 100,000 population). https://www.who.int/data/gho/ indicator-­metadata-­registry/imr-­details/4427 Zsigmond, B., Breathnach, A. S., Mensah, A., & Ladhani, S. N. (2022). Very low rates of severe Covid-19 in children hospitalized with confirmed Sars-CoV-2 infection in London, England. Journal of Infection, 85(1), 90–122. https://doi.org/10.1016/j.jinf.2022.04.020

Chapter 5

Public Choice Theory: An Explanation of the Pandemic Policy Responses

Public choice theory suggests that the pandemic policy responses were in fact the result of politicians’ and bureaucrats’ ambition to pursue their own interest. The main target of politicians during the Covid-19 pandemic was the one that politicians typically aim at, i.e., to maximize votes. Lockdowns, mass vaccination, and vaccine passports are largely explained by the vote maximizing premise. Bureaucrats such as scientists who work for the government but who do not appear as candidates in the elections also engaged in utility maximizing during the Covid-19 pandemic by pursuing their own goals which include, among others, increased popularity and willingness to establish their reputation. A budget maximizing analysis is used to illustrate how an ever-increasing budget satisfies nearly all groups involved in the pandemic, i.e., politicians, bureaucrats, and voters. Findings from polls and election results verify the public choice analysis.

5.1 Can Voters Ever Be Public-Interested Agents? We have seen that the data that were used to justify lockdowns were mistaken and that only people above 70 face a high risk of dying from Covid-19. We have also seen that lockdowns do not make older people better off (they do not save them from the virus), and it appears they make them worse off (lockdowns likely expose them to the virus). We moreover saw the YLL and the very high HRs due to reduced income and unemployment as well as the concerning effects of the  constantly increasing inflation rates. Moreover, isolation also impacts on well-being in several respects. Thus, lockdowns do make young and middle-aged people worse off. These results refute all the assumptions upon which the public-interest theorizing on lockdowns was based. In spite of the fact that lockdowns were ineffective and damaging, a group of people ended up better off after their implementation, namely politicians and © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 P. Karadimas, The Covid-19 Pandemic, Studies in Public Choice 42, https://doi.org/10.1007/978-3-031-24967-9_5

97

98

5  Public Choice Theory: An Explanation of the Pandemic Policy Responses

bureaucrats in regions that enforced such measures. Down’s insight seems to fit the case. As he stated, the relationship between policies and the expectation to win elections is not ambivalent, and it certainly is not the case that the latter stems from the effectiveness of the former. The converse is in fact what happens. That is, politicians (or parties, as he puts it) formulate policies in order to win elections and do not win elections because they implemented cost-effective policies (Downs, 1957, 28). By slightly tweaking it and reapplying it to the covid crisis, the claim is that politicians implemented lockdowns (and lockdown-like interventions such as mask mandates) and other sweeping interventions, such as mass vaccination programs, so that they will win elections rather than they won elections prior to the pandemic in order to implement policies akin to lockdowns; and of course, they do not expect to win elections (as we will see, some of them have already done so) because lockdowns are beneficial to a sizeable fraction of the population, but because the public thinks that lockdowns and the like were the optimal policy. While the vote-maximizing premise, which will be developed in this chapter, explains very well the policy responses and the actions of politicians during the pandemic, it is not enough to explain the actions of unelected bureaucrats, such as members of scientific committees that supposedly followed science and offered policy proposals that are backed up by sound evidence. Bureaucrats also engage in utility maximizing, but since they do not appear as candidates in the elections, they respond to different incentives and constraints than the ones elected politicians do. Their main goal is to increase personal utility (which translates into “power” in this context), instead of working as public servants and constantly estimating which decisions best serve the so-called common good. An ever-increasing budget is what made lockdowns and mass vaccination possible and thus helped politicians and bureaucrats attain most of their goals while satisfied the public who felt protected from the virus while receiving handouts to stay at home. These two propositions of public choice economics, i.e., the vote maximizing and the budget maximizing premise are united and both provide us with a comprehensive scientific explanation of the pandemic decision-making. To have a complete account of the public choice view of the pandemic, we need to establish not only that politicians and bureaucrats are utility maximizers but also that voters are. If voters are public-interested individuals, then politicians would not have been utility maximizers for a very long time, and especially during the pandemic, whereby the policies that were implemented were at glaring odds with scientific evidence and caused remarkable damage. Moreover, saying that voters are public-interested agents would have been inconsistent with the main premises of public choice theory; since politicians and voters are individuals, the motivations need to be the same and only the incentives and constraints to which they respond need to change. The premise that individuals act so that they will maximize their utility in all branches of life is well-documented, and one can hardly object its robust empirical grounding. Thus, people act as rational agents during voting as well and thereby hinges the “rational voter hypothesis” which was first developed by Downs (1957). However, it has led to a paradox that public choice accounts cannot leave

5.1  Can Voters Ever Be Public-Interested Agents?

99

unaddressed, namely, to explain why individuals vote in the first place—the socalled “voting paradox.” Indeed, it is not at all clear that it is rational to participate into voting procedures. As public choice theorists argue, the probability of one person’s vote to be the decisive one is lower than the probability of being hit by a car while going or leaving the polls (Goodin & Roberts, 1975), so if voters are rational individuals the rational action is not to vote. The rationality assumption thus seems to flounder which could have serious implications for our analysis and for any public choice analysis. So we need to answer why people vote and why they vote as they do and explore how rationality can be retained. In my view, the voting paradox issue is tightly linked to whether voters are public-interested or self-interested individuals. By showing that the latter is the case, we save the rationality assumption, while simultaneously explaining both why people vote in the first place and why they vote as they do. On the contrary, if we assume that voters are public-interested individuals, we either need to abandon the rationality assumption or to explain only why they go to the voting booth without accounting for why they vote as they do or conversely to explain only why they vote as they do without explaining why they vote in the first place. That is, if we posit that public-interested people go to the voting booth and vote for the candidate that offers policies that make society better off, we go back to square one for it is irrational to go to the voting booth. And if it is irrational to vote, it could follow that only irrational people are the ones who vote, and hence, if that is so, we need to explain why people flock to polls on the Election Day, and it is an exaggeration to dismiss all of them as being irrational. But even if we do so and say that anyone who votes is irrational, the problem is far from solved, for if so many people make an irrational choice and go to the voting center, then it can hardly be argued that so many people that made an irrational choice initially can now make such a rational action by casting a vote that promotes public good and takes into account all the possible tradeoffs that appear in a society. A possible way to insist on the public interest theory of voters and reconcile it with the rationality assumption is to appeal to a notion of “civic duty.” This approach discriminates between the participation in the elections and the decisiveness of the vote by claiming that voters vote due to a sense of civic duty from which they derive pleasure are not interested in whether their vote will be the decisive one.1 However, we have no clue as for how people cast their vote and it cannot be argued that people vote for the “common good” or that their civic duty drives them to vote the best policy, for since they have realized that the decisiveness of the vote does not matter, and they vote to satisfy their “civic duty,” it is hard to say that they actually act as public-interested

 The “civic duty” hypothesis was initially introduced by public choice theorists in an attempt to rescue rationality from the voting paradox and to claim that the act of voting brings benefits to the voter and thus explains why people vote. However, an important public choice theorist, Dennis Mueller, seems to allow that the civic duty assumption is compatible with a public interest approach to voters’ actions as well (Mueller, 2003, 306). Thus, it is, I think, important to comment at least briefly, on why the notion of civic duty per se does not suffice to enhance public interest theory of voters’ behavior. 1

100

5  Public Choice Theory: An Explanation of the Pandemic Policy Responses

individuals because a public-interested voter needs to care about the decisiveness of their vote and not about the possibility of benefiting personally from the act of voting.2 Rational irrationality, its microfoundations of which we discussed in Chap. 3, offers the theoretical framework to account for both why people vote and why they vote as they do, without relaxing the rationality assumption and without making concessions to the public-interest theory. Caplan argues that, at some level, people realize that their vote is not the decisive one (Caplan, 2007). However, if voting is costless at a personal level,3 as rational irrationality suggests, and offers individuals an uncommon4 benefit, namely to act on ones’ own beliefs, this makes the act of voting rational. In this framework, the main purpose of the voter is not to cast the decisive vote, but instead to indulge in acting on their own biases, which offers a strong incentive to participate in the voting procedure. Since they vote in order to satisfy their own biases, they vote for policies that are in accordance with these beliefs. People often hold beliefs that scientists consider them to be wrong (though this is not necessarily the case) and cast their votes in favor of policies that are in agreement with the voters’ established beliefs. Most importantly, they have no incentive to change them since they do not act on them during their everyday lives, and thus, they do not suffer the cost of acting constantly on their own beliefs. This in turn explains why people choose bad policies; once the policies on offer satisfy their established beliefs and if their beliefs are unrevised and (typically) scientifically mistaken, it is likely that their votes will favor policies that are bad. Indeed, to discriminate between the act of voting and the votes people cast does not suggest that one of the two is of minor importance and is in fact the key to solve the voting paradox and make the case for voters as self-interested individuals. The notion of “civic duty” therefore that can be theoretically used within the framework of public interest theory appears to be more compatible with the analysis of voters as self-­ interested persons. Self-interested individuals vote in order to avoid blame for breaking the civic duty to participate in the elections and they benefit from this. Moreover, when they arrive at the voting booth, they vote for policies on a par with their fixed beliefs while suffering virtually nil cost. Therefore, Caplan’s theory is  Other attempts to save the rational voter hypothesis have been proposed by Ferejohn and Fiorina (1974, 525), who describe the voter as a rational agent who chooses the option that minimizes regret while maximizing the gains. They name it “the mini-max strategy.” They argue that if the gains of having one’s preferred candidate elected are double the costs of voting, then the mini-max strategy is to vote and vote for one’s preferred candidate. However, this approach has been widely criticized by economists who question its applicability either when it comes to voting or in real-life situations (Goodin & Roberts, 1975; Mayer & Good, 1975). 3  While this is not literally true, for it takes some time to go to the polls and perhaps it takes also some time to queue in line and this amount of time could have been spent differently in case one chooses not to vote, it is fair to claim that the cost is so minuscule that we can, loosely speaking, albeit entirely justifiably, say that it is near to zero, and that thus the act of voting is costless. 4  It is uncommon for as we saw in Chap. 3, when one acts on their established beliefs during their everyday actions, this carries remarkable cost and so most people avoid it. The only exception is the act of voting whereby the cost is near zero. 2

5.2  Voters’ Ideal Point: A Conglomerate of Expected Utilities

101

wider in scope and can be used to explain first why people vote, why they vote as they do and simultaneously can help us establish why voters are self-interested persons and not public-interested ones. The description of voters as public-interested individuals does not manage to overcome the voting paradox and unless this is solved from a public-interest perspective, it can hardly be argued that there are solid grounds to consider voters as public-interested agents. Voters as self-interested individuals appears to be the only scientifically plausible description of the majority of the people who vote.

5.2 Voters’ Ideal Point: A Conglomerate of Expected Utilities It appears therefore that, as Downs first noted, voters form utility expectations and they expect politicians to fulfill them. Tullock elaborated on the idea (Tullock, 1967) and so did Riker and Ordeshook (1973). This seems to involve two strands of voters’ expected utility; the one is related to the fact that people expect politicians to offer policies that would satisfy their preconceived biases and the second is to offer policies that would satisfy each voter’s or each groups of voters’ particular interests. Therefore, the politician who maximizes most is the one who offers Pareto optimal policies, i.e., policies that manage to reconcile the two: to satisfy to a great extend widely held beliefs in the society, while simultaneously to satisfy the interests of particular groups. If a group, say schoolteachers, receives extra money from the government, other pressure groups (say workers at the public transportation) would take exception to that even though they could generally value highly expanding fiscal policies. They could rebel in such a case, because they could think that their personal interests are not served and that the government could have given money to other groups, too. Thus, those who would be disappointed by this policy may constitute a larger segment of the society than those who would support it and the politicians who pursued this policy may end up worse off despite the fact that they enforced a policy which was in accordance with widely held beliefs. In such a case, the voters do not revise their belief but they would vote for a candidate that would satisfy their bias, for example, someone who would promise more public spending than the one that took place. On the contrary, if a politician promises to construct railways at some remote areas of a country, the majority of people would back this idea despite the fact that the vast majority would participate in the costs of the project through taxation without receiving any benefit and only a tiny fraction of the population would become better off (those who live in the remote areas). The reason they would support it even though it makes them worse off (unless they pay in taxes an amount of money that is less than the benefit they perceive, which scarcely happens in real life) is that it is socially desirable to construct railways in remote areas and in contrast to the teachers’ example, it is most likely that they would not expect government to help themselves in a similar way. It moreover fits biases people typically hold according to which social welfare services are optimal. At the same time, those who would directly benefit from the project would wholeheartedly support it

102

5  Public Choice Theory: An Explanation of the Pandemic Policy Responses

because it serves their own interests despite the fact that it poses a cost to others. We can assume that if elections are about to follow in such a scenario, the politician who implemented the railways in these areas would gain votes both from the general public since this policy was in accordance with people’s beliefs and from the chunk of the population that will use the railways. These examples are imaginary and their only purpose is to showcase that the biases and the fixed beliefs society holds by and large and the particular interests voters or groups of voters could have are not always identical, and politicians need to successfully reconcile them to serve their interests. When the two are combined, the voters’ ideal point is formed. While there is no clear threshold for this, the general rule could be that voters expect policies that satisfy widely accepted biases which even though come at some cost on personal interests, the groups who suffer the cost do not outnumber those that either benefit or are neutral with respect to costs but nevertheless see their biases being satisfied via the policies in place. To elaborate a bit on how the voters’ ideal point is formed, we can offer a distinction between what happens during “normal times” and what happens during “abnormal times.” By and large, normal times can be considered the years after the Second World War and prior to the pandemic and abnormal times episodes that upend societies in a very short period of time, like the Second World War and the Covid-19 pandemic.5 In normal times, the voters’ ideal point expects policies that satisfy their beliefs at some minimum personal cost, which typically goes unnoticed, especially if the policy in place is socially desirable (as the example with the railways implies). During abnormal times, the voters’ ideal point involves satisfying their biases even if the personal cost is huge. The reason is that during abnormal times voters believe that they benefit when policies in accordance with their biases are implemented. So, most people typically support wars, as, for example, happened with the Second World War (Larson, 1996), even though wars seriously harm the economy and thus the majority of the population. Therefore, the main patterns are the same either in normal or in abnormal times and what is altered in each case is the magnitude of the voters’ expectations and hence typically the magnitude of the force the policies are implemented. Widely held beliefs in a society seem to permeate across different social groups of voters. In the example on the construction of railways, I relied primarily on the intuition that people typically expect the governments to take action instead of not intervening and that the lengths that they expect the governments to intervene is hugely dependent on whether we are in times of crises or not and certainly not on whether the type of intervention is evidence-based or on whether it makes the majority better off. However, as stated briefly above, the system of beliefs that guide voters’ expected utilities need not necessarily be scientifically mistaken. People can have fixed beliefs that are scientifically debatable or even accurate. For example,  As this book tries to demonstrate, it is not the pandemic per se that wreaked havoc but the fact that mass hysteria stroke and that politicians implemented devastating policies in response to the widespread panic. In any case, it can be described as a deep crisis, and hence it is an abnormal state of affairs which affects the way the two strands of voters’ expected utilities correlate. 5

5.2  Voters’ Ideal Point: A Conglomerate of Expected Utilities

103

nearly everyone believes that currency is needed for a functioning economy which is a belief that is not a bias that scientists deem it as problematic or wrong, but it is a well-established premise in economics for as Menger demonstrated long ago, money is a good that people accept as the most marketable one (Menger, 1892). Thus, a policy that would suggest abandonment of the institution of currency in favor of a moneyless economy would have been nowhere near what voters expect and it would moreover contradict sound economic postulates. Value judgments are also part of the story when it comes to voters’ beliefs and they are certainly part of the voters’ ideal point and hence of the policies that are followed. Alesina and Angeletos provide an analysis of multiple equilibria economies whereby fixed beliefs different societies hold about fairness and equality determine different policies with respect to redistribution and to the expansion of welfare state. The higher societies rank these values, as it happens in Europe, the more the welfare state is expanded and, on the contrary, the lesser they value them, as in the USA, the lesser it expands. If voters believe that free markets do not reward effort, they support a larger welfare system; but a larger welfare system undermines effort and thus the voters’ beliefs become self-fulfilling and similarly, if they believe that everyone needs to enjoy the fruits of their labor, they will vote for low taxes and hence lead the economy to an equilibrium whereby effort is high, thus making the beliefs self-­ fulfilling too (Alesina & Angeletos, 2005). Their analysis shows that fixed beliefs in different regions of the world shape policy making by pushing politicians to respond to different incentives and constraints in different parts of the world. In the Covid-19 pandemic, whereby mass hysteria dominated the world, the fixed belief in all societies was nearly unique, i.e., all people are likely to die of the virus and thus the policies that need to be put in place should aim at protecting all people. Edgeworth’s box can be used to illustrate the correlation between the strands of voters’ expected utilities we described in this section and how they correlate with the policies politicians offer. In the framework of the Covid-19 pandemic, there is one such attempt available in the literature by Zweifel whose contractarian approach compares how people used to handle risk prior to the pandemic and how this radically changed when the pandemic broke out. Before the pandemic started off, people were willing to handle everyday risk on their own, but during the first wave of the pandemic, they relinquished this attitude and expected politicians to make the relevant risk-assessments. In fact, they exchanged freedom for safety (Zweifel, 2020). Until early 2020 (during normal times, according to the terminology used here) people were unwilling to give up their everyday lives. However, during the pandemic (the abnormal times) the vast majority of the population was convinced that they face an unprecedented health crisis, even though this was not backed up by the data. Therefore, their expected utility was to find a way to avoid the virus. Hence, their ideal point shifted and in the formation of it, the biases dominated over personal costs while even just prior to the pandemic loss of basic freedoms and a shutdown of the economy would have been considered as an unpleasant (to say the least) cost.

104

5  Public Choice Theory: An Explanation of the Pandemic Policy Responses

5.3 Politicians as Vote Maximizers Mueller’s probabilistic-voting analysis helps illustrate politicians’ decision-making during the Covid-19 crisis. Mueller shows that the number of expected votes is the sum of the probabilities that each voter will vote for the candidate. Voters choose the candidate who is closest to their ideal point, and politicians tend to adjust their decisions accordingly (Mueller, 2003, 252). Since tackling Covid-19 at all costs became the foremost objective for most of the voters, policies that would meet this demand have become their ideal point of reference. Voters’ ideal points were solidified by herd behavior whereby one government followed the example set by others. This herd behavior of policymakers was first observed in the European Union, where Italy followed the example of China6 and Iran, and then other countries (including the UK) followed the example of Italy (Sebhatu et al., 2020). The public in countries that had not already locked down therefore were expecting politicians to take sweeping measures as the only appropriate approach.7 It appears that there are three main public policy alternatives in times of pandemic: First, take no action and let the virus circulate. Second, take moderate action, such as the “focused protection” plan sketched above. Third, implement lockdown policies. If mass hysteria prevails, as it happened during the Covid-19 pandemic, then the more active the politicians are, the more they maximize their utility. We can represent the impact each one of the three alternatives has on politicians’ utility maximization, by using the following diagram (Fig. 5.1). Figure 5.1 shows the votes politicians would gain in each case. The public’s concern level is represented by the horizontal line (x-axis) and the number of votes politicians gain by the vertical line (y-axis). Politicians pick their policies based on how many votes they will receive. As it is shown in Fig. 5.1, point V represents a number of votes lower than V′, which in turn represents a number of votes that are fewer than V″. Because the public’s main concern is Covid-19 and their ideal point is based on the belief that all face serious risk from the virus and all are in need of equal protection (EP), politicians gain few votes by taking no action (NA). In that scenario, the politician receives only V votes because the public believes the politician cares little for public health—only a handful of voters who have other priorities, such as issues germane to the economic impact of the measures or to school

 Leaders in totalitarian regimes also act as self-interested individuals. The head of the CCP in China responded forcefully and enforced lockdowns not to maximize votes, but to show its citizens and the world that authoritarian decision-making does what it takes to address the problem. This convinced Chinese people that the regime is to be trusted and it simultaneously set a paradigm for the world. Thus, panicked people in the west demanded forceful china-like action and their politicians gave them what they asked for. 7  The war-metaphors, frequently expressed by politicians, were part of the vote maximizing procedure and of the willingness of politicians to meet the public’s ideal point. Among others, Trump, Macron, Johnson, presented themselves as individuals who are willing to save the population from a war-like hardship. This in turn instilled more fear in the population and hence it further cemented the public’s ideal point. 6

105

5.3  Politicians as Vote Maximizers

V″

LD

Number of Votes

V″minus

RL

V′

V

MA

NA CM

CM′

EP*

EP

Voters’ Preferences Fig. 5.1  Politicians’ utility maximization. (Karadimas, 2022)

closures, or concerns about excessive governmental intervention vote for the politician. Since even the long-lasting lockdowns faced little to no backlash, I categorize possible pushbacks as concerns of the minority—that is, of the voters who would prefer no action or moderate action (CM and CM′, respectively). So, if politicians took moderate action (MA), they would not receive many votes, though they would have been better off than by taking no action. Moderate action falls short of meeting the public’s ideal point, because it proposes targeted measures for the elderly and not for the entire population and thus it reaches point V′. By implementing lockdowns (LD) and offering equal protection to all, they are on a par with the public’s ideal point and they remarkably maximize their utility (V″) (Karadimas 2022). In the same mode, we can represent the levels of severity of lockdown policies and the respective gains for politicians in each case. Let the first lockdown level include business shutdowns, mandatory masking indoors and work-from-home mandates while schools remain open and no stay-at-home mandates are imposed. Name it Light Lockdown (LL). Let the second level include school closures, business shutdowns, and work-from-home mandates, but again no curfews or twentyfour-hour stay-at-home mandates are ordered. Name it Moderate Lockdown (ML). The third level encompasses all the contents of a lockdown strategy: schools and businesses are closed, almost everybody works from home, and twenty-four-­hour stay-at-home orders are imposed. Call it Strict Lockdown (SL). All levels of lockdowns include mask mandates. Figure 5.2 represents the utility politicians would gain in each of these cases. The x-axis shows voters’ preferences under such circumstances, and the y-axis represents the number of votes politicians would gain (as in Fig.  5.1, V