The Risks of COVID-19

When Mark and I started this project, we had planned to steer clear of politics, wishing to avoid the rancor in that subject. But it's hard to discuss risk without addressing the elephant in the room: COVID-19. After all, it has been the longest, and deadliest, epidemic in more than a century, and its course in various countries is inextricably linked to risk attitudes.

On the surface, you would think that a disease like COVID would follow “normal” risk parameters, and in isolation it probably does. Yet because this is not purely an individual, but rather a societal issue, COVID can’t be viewed in isolation.

As Mark writes in his post on helicopter parenting,  people generally overweight unknowns in their risk assessment, and a disease should certainly qualify. But this one is different in large part because COVID and the measures governments have taken to combat it have had an oversized effect our lives compared to other diseases. In the years to come, I am sure the pandemic will be the subject of many a doctoral thesis not only in epidemiology, but also in sociology, politology and lots of other –ologies, potentially even theology (eschatology, anyone?). So for today, a few thoughts in no particular order and with no claim to completeness.

Science is confusing

Unfortunately, when there is a complex problem, most people prefer simple answers. Our modern media environment exacerbates that: (bad) news sells better than boring facts.  For those of us old enough to read or have read newspapers (I still have a subscription, living in France, to Le Monde), the fact you got it daily and at least browsed made you more open to reading things slightly outside your normal knowledge zone. But when headlines and clicks are all that count, it becomes harder and harder to educate the public in things like COVID.

Add to that the fact that real experts are often boring and suck at explanations in laymen’s terms, and that the scientific method is slow and fraught with uncertainty – that’s the point, really, to gather data and advance knowledge bit by bit. Unclear terminology and apples-to-oranges statistics comparisons don’t help either. Country x has more cases of the disease than countries y and z. But x is larger, and y has tested more – is z really better? How many people have died of COVID? Or do they actually die with COVID, not of ?  In one sense it does not make a difference: all that matters is how many people wouldn’t have died now if they hadn’t contracted COVID.[1] But this confusion makes planning and remediation more difficult, and undermines public trust in science and scientists. Which brings me to:

Who do we believe?

Echo chambers and epistemic bubbles are nothing new.[2] We’ve always read the newspapers with our own partisan lean, mingled with people “on the same wavelength”.  Facts contrary to our own thinking always induced cognitive dissonance. In the pre-digital world, however, the bubble was seldom impermeable – it was hard to escape opposing opinions and facts. The internet, and specifically social media, has changed that quite a bit – just hop on Reddit and pick a few subs at random to get some fine examples of echo chambers at work.

Some time ago, Michael Shermer published a book called “Why People Believe Weird Things.” An interesting read and worth reviewing in the context of COVID; the one short takeaway of the book for this post is that believing weird things is an evolutionary maladaption. Let me explain:

In statistics we can make two basic types of error, type I and type II. (And these can be translated to psychology). A type I error is a “false positive”, i.e. finding something when nothing is there. A type II error is the opposite, finding nothing when something is there.

Type I errors are maladaptive.  Imagine one of our ancestors on the African savannah. The grass is moving. Is it the wind or a predator? If it’s the wind but he falsely believed it to be a lion and runs away, he has committed a type I error, and will live to procreate (and never learn of his mistake). If it’s a lion but he doesn’t believe it, i.e. a type II error, he learns of his mistake very quickly, and also doesn’t hand down his genes. Over time, ‘type I error genes’ will vastly outnumber those for type II.

Why does this matter in this context? After all, not believing in a disease that exists is a type II error. That brings me back to the echo chambers, and the question who we believe. By their very nature, epistemic bubbles have an incomplete view of the real world, and its ‘experts’ often aren’t – a fertile breeding ground for half-truths, untruths and conspiracy theories, the belief of which is a typical type I error. So it’s no surprise that the public’s risk assessment of COVID is all over the place. One chapter in Shermer’s book is title “Why SMART people believe weird things”. The answer: for the same reasons everybody else does, but because they are smart they are better at rationalizing their beliefs. And politicians are also (generally) smart people and often live in the same bubbles. So, next:

It’s all about politics

When echo chambers become increasingly impermeable, an increase in political partisanship as we are seeing in many countries, not just the US, is a natural (inevitable?) outcome. Suddenly everything, including a deadly pandemic, becomes political and is seen through a partisan lens. Because ‘they’ are saying x, it must be wrong, so ‘we’ have to say not x, facts be damned. I won’t argue whether the consequences of polarization are symmetrical in general, but in the case of COVID and its risks it’s clearly not. Even if wearing a mask were useless to stop the disease from spreading, wearing it has no downside apart from some temporary personal discomfort. (No, you don’t suffocate when wearing a mask all day – otherwise hospitals would be full of dead surgeons.) The other direction – it’s useful yet people don’t wear them – has vastly different and negative consequences.

Why is mask wearing even a question of politics? At the foundation, it goes back to the question of culture and its consequences I discussed in an earlier post: What is more important, my liberty or your rights? Is it a sign of weakness and do I want to be seen as weak (as a culture)? And, what do the leaders in my echo chamber say?

A case has been made that countries with female leaders got through the pandemic much better so far. German chancellor Angela Merkel’s COVID explainer video rightly went viral, though her being a trained scientist may have helped.  Cause and effect are less than clear: are female leaders really better at this, or are countries that are emancipated enough (whatever that means in this context) to elect female leaders more likely to follow rules that are aimed at ameliorating an epidemic? While I personally believe it’s a combination of both, there are a few counterexamples: US states that have female governors that haven’t done well. The UK had two female prime ministers in the past, yet are currently not exactly a shining example of how to handle COVID.

It’s the economy, stupid

Last but not least, beneath this all is the fact that the pandemic itself and the additional effects of countermeasures like physical distancing and lockdowns are having a painful effect on the economy and on many people’s lives. Too many people are facing the choice of a potentially becoming ill versus certainly not being able to pay rent. Again, this veers into the political – people in countries with a functioning societal safety net are in a much better position to follow their risk assessment than to have their choice dictated by necessity.

But no matter what your personal view on COVID and its risks are, one thing is pretty certain: the world will not look that same afterward.

Notes:

[1] In many cases, people do not die directly from COVID, but rather because the virus either exacerbated already existing other health conditions, or introduced and complicated new ones. Technically people don’t die of AIDS either, but they wouldn’t die of other conditions like pneumonia if AIDS hadn’t ravaged their immune system.

[2] Both are social structures which are one-sided regarding opinions and often even facts. In epistemic bubbles, that exclusion can  be accidental, whereas in echo chambers it is intentional and active.