Horse to Water
What level of risk justifies extraordinary government decrees?
Our fear of close human contact is well beyond the real measurable risk.
Government policies are focused on the wrong metric (declining cases) and are far too strict.
We can only hope to return to normality by fully re-opening our societies.
Confidence Trick
The coronavirus pandemic is actually not the social problem of the moment.
It is certainly barreling ahead and many places have yet to reach their peak contagion, notably in South Asia and South America.
However, for big parts of the planet, including most of East Asia, North America and Europe, we seem to be well past the worst days of March.
And while the number of world-wide positive cases goes ever higher, in large part due to increased testing, the number of new deaths is still one-third below its peak level.
No, the most pressing problem for us all is the social response to the pandemic and, in particular, the widespread fear of mixing with other people in close quarters.
This exaggerated perception of risk is lasting well past the point when the data and the reality of risk are receding in most places. It is more dangerous than the pandemic itself because we will have no lasting economic recovery, or even peace of mind, until people go about their lives in a more normal way.
It is a collective psychological problem and so can only be resolved in our minds. And the way to do that is to instill confidence in people, such that they stop thinking about health risks in every part of their daily activities and start implicitly trusting strangers and public spaces again.
Setting Risk
Do not rely on consumer confidence or business sentiment to enlighten your understanding of social confidence and the possible ways back to normality.
Those concepts are far too narrow. They cover mere economics, when the theme of our time is a full-on humanity-wide social crisis.
The bigger issue to focus on is trust in our fellow man (and woman) and our limited willingness to move in close and rub shoulders with each other.
The desire to touch and hug and engage in close proximity is our natural default position. We desperately want to be with each other and we want to interact in a way that creates harmony and community.
That is why today’s unnormal environment is so alien and can only be transitory, a rough moment in lives otherwise massively filled with close contact with other people.
So, what is holding us back, when what we really want is to (almost) rush into each other’s arms?
Fear is the main factor. And, specifically, fear of an invisible foe, a coronavirus that is potentially carried and transmitted in every breath we breathe.
So long as there is even one Covid-19 case reported anywhere in the world (and we all know that case will be loudly reported), there will be a significant fraction of people who will avoid close human contact.
This behavior is absolutely excessive and illogical, given the physics and likelihood of transmission, but it is inevitable since exaggerated perceptions of risk are completely dominating the reality of risk.
The only way to assuage this fear is to wean people off their isolation and deliberately expose them to others. With repetition and the observation that renewed human contact is safe, a hesitant start will eventually turn into a habit and finally an un-self-conscious daily practice.
To do this, we need to accelerate the re-opening of our societies in all locations where the pandemic has peaked and then moderated to stable levels.
The objective is not zero cases, hospitalizations, or deaths, since every activity in life comes with some risk. Rather, for this disease, we need to aim for a low and non-accelerating level of contagion that circulates in low risk populations (those who are under 65 years old and with no relevant co-morbidities).
At a risk setting commensurate with other infectious diseases like the flu, or other social ills like suicide, workplace accidents, or auto collisions, we can then gradually expect to see a social response that is proportionate to a normal risk.
Defining Contagion
The first question to consider in setting a risk target is what to measure.
We want to track contagion but all of the real-world measures are mere proxies for the brief human interaction that leads to transmission and infection. The real risk is the probability of contagion combined with the subsequent probability of severe illness.
There are a number of candidates to measure this risk:
ILI (influenza-like illness) data are the timeliest indicator but are not widely available around the world and do not indicate severity or distinguish between viral diseases.
Positive cases data, usually detected after the person becomes symptomatic, ramp up with the amount of testing and so may simply reveal contagion that already exists. They also ignore severity.
Hospitalization data are reasonably accurate but are not widely publicized. These data are also becoming harder to interpret because most non-Covid patients are now being tested and hospital activity is rising back to normal capacity levels.
Finally, mortality numbers are the least likely to be distorted and are what we fear most. Further, we can easily compare death statistics across other risk categories. The main drawback is that deaths occur on average three to four weeks or more after contagion.
Given these choices, I would opt for death statistics as the best indicator of contagion and risk, even given the time lags. The data are widely available and tend to trend in a stable way.
Who’s at Risk?
No one has to guess at what a low level of contagion might be. We can use math to come up with reasonable scenarios that match other risks in life, where we do not engage in a prolonged social panic as the primary response to the threat.
For example, looking at the US, we see that there are 8,678 deaths per million people each year. That is a rate of just under 0.9%. The leading causes of non-chronic illness are accidents (at a rate of 0.052%), flu and pneumonia (at 0.017%), and suicide (at 0.014%).
Right now, based on the latest week’s data and assuming that rate for the next year, the US death rate from the coronavirus is running at 0.07%, almost one-third higher than the accidents rate alone. So, on the surface, it looks like the current pandemic is considerably more dangerous than other threats.
This, however, ignores two basic issues: the viral death rate is skewed too high and not everyone has the same risk exposure.
The skew comes from categorization bias, where finding or assuming the existence of the virus in the deceased is taken as the primary cause of death, whether true or not. This error will become more profound this winter when we head back into flu season and attribute some flu deaths to the coronavirus.
Secondly, about half of all deaths have occurred in long term care homes, an assembly of extremely vulnerable people in a dangerous location who do not at all represent the populace at large.
Finally, the death rate is also exaggerated over the course of a year because the virus (so far) has a highly concentrated effect, accelerating in the early days and then usually moving well below the peak to a lower stable level. It is also novel and its introduction to the human biosphere is likelier to be more damaging now than in coming years.
The issue of dissimilar risk exposure from one person to another is more profound.
For young people in the US under age 25, the viral death risk is negligible, at a rate of less than 2 deaths per million. Accidental death risk is 90 times higher, suicides are 40 times higher, and flu and pneumonia risk is twice as high.
For working age people to age 55, viral mortality is more serious, with almost 90 deaths per million, near death rates for diabetes, breast cancer, or cirrhosis, and focused on the older people in that cohort. Still, accidental deaths are 6 times higher and suicides are twice as high, while flu and pneumonia deaths are 3 times lower.
It is only after age 65 that deaths from Covid-19 exceed the other three non-chronic leading causes of death, with lethality rising sharply and tragically with age.
And that suggests a radically different public policy from the one most countries and states have followed: quarantine or restrictions for the elderly and the sick and relative social freedom for the rest of us.
When to Panic
But where is the threshold to judge granting or restricting the freedom to mingle with others? At what point do we as a society take extraordinary measures to combat a truly serious threat to our health?
The experience of this pandemic is that political leaders around the world have generally done a poor job of judging that critical moment.
They did too little at the start of the pandemic and then later rushed into a social lockdown at the same time, over the course of two weeks in mid-March, irrespective of local pandemic conditions. This suggests a policy panic where there was extreme pressure to respond to growing public fears.
Now, we have partial re-openings that are tepid and delayed, and with such restrictions that many normal daily activities are forbidden or marked by needless discomfort and inconvenience. A renewed rise in positive cases in some jurisdictions is even threatening to reverse the opening process.
This multi-directional pendulum swing from lackadaisical to draconian to hesitant policies has been based on a mishmash of poor information sources: epidemiological forecasting models that grossly exaggerated the threat, public health officials who ignored economic and social tradeoffs, and just plain ignorance of how our modern tightly connected society works.
So, instead of putting our finger into the wind, or myopically taking local conditions as the norm and seeking to reduce them, we should use the risk assessment approach outlined above to set trigger levels for policy action that are related to real risks.
It is important to note that this cannot be done on our present policy course, where governments only grudgingly open up social interactions in response to a declining trend in positive cases and refuse to go further until cases drop to zero.
Given that an effective vaccine may only come in a year or two (or more probably never), the only other way that the pandemic will stabilize is through widespread immunity, which by definition means more contagion, the opposite of current government objectives.
We therefore need a wholesale change of perspective, so that each new infection is seen as moving us one positive step closer to the end of this pandemic. The trick is to isolate vulnerable populations to ensure that the inevitable spread produces overwhelmingly mild or no symptoms and minimizes the number of hospitalizations and deaths.
We should therefore switch course and move the public health policy benchmark from tracking declining cases to actually measuring and publicizing the real risks of contagion.
By doing so, we will be in sync with the dynamics of the pandemic, since it will have its way with us, no matter how fervently we wish for its early demise.
Triggered
So, finally, let’s consider the triggers for instigating public health controls based on the real risks of contagion.
Again, using the US as an example, we know that the health risk of key non-chronic events (accidents, suicides, flu and pneumonia) is equivalent to a death rate of just over 0.08%. For those under age 65, it is less than 0.06%.
These threats do not call forth economic lockdowns or stay-at-home orders or any other coercive measures to ensure public compliance.
Rather, we have a set of socially accepted rules and practices that moderate the risk, ranging from workplace regulation and car design to counselling and vaccine programs. Given this backdrop, we go about our daily lives relatively free of fear.
So, our first real risk trigger should call forth only normal regulation as a public policy response, since it is equal to the existing non-chronic risks that operate that way.
Right now, only 12 US states (in the North East, Mid-West, South, and South West) are above that level, with annualized death rates in excess of 0.08%, and 5 others are just under that level. By contrast, 22 states have rates at less than half of that level.
There is thus room in the US for a variety of policies to reflect a variety of conditions. Interestingly, given current media concern, big states like Florida and Texas are not in the higher risk group, while Illinois, Pennsylvania, and New Jersey are included.
Applying this policy guidance around the world, we can see that there are only 9 countries in the same higher risk situation (and all of them except Sweden are in the Balkans or Latin America). By contrast, over 180 countries have annualized current death rates that are less than half of the 0.08% level.
This suggests that current government public health policy around the world is far too tight.
Going further, we can ask what is the risk level at which we should instigate society-wide controls, since they have such dire impacts on the economy, our freedoms, and our social relations.
Taking an arbitrary higher risk level of twice the non-chronic events, which can be justified by the extraordinary conditions of a ragging pandemic, we find that only five states (Rhode Island and Massachusetts above the level and DC, Illinois, and New Jersey just below) and four countries (Chile, Brazil, and Peru above the level and Mexico just below) qualify for the strictest public policy measures.
Again, we can see that governments are being too heavy handed across the world since every country is under rather strict controls at the moment. Much faster and preferably immediate re-openings of the economy and public spaces are obvious conclusions that come from this analysis.
We can play with other variations on this theme but the answer will be the same: the real risk of the coronavirus, as measured by mortality rates, is too low to justify the current heavy-handed public policies.
Will governments respond by taking their foot off the brakes?
Will the media publicize the relatively low death rates and the relatively tight public policies?
Will people eventually relax as they renew their social relations?
Sadly and probably not, as you can lead a horse to water …
