On coronavirus and a sick giant

I’ve long been a fan of Tim Urban’s brilliant blog Wait Buy Why. His recent post A Sick Giant is the latest in a long series on how human individuals and societies process information and make decisions. While Tim published this post in January, before coronavirus entered the world stage, his theories about why recent social and technological changes have lead to unprecedented political polarization can also, in my opinion, explain the world’s reaction to this pandemic.

The question that I’ve found most fascinating related to coronavirus is why, in 2020, did the world almost universally ‘lockdown’ to slow the spread of a disease for the first time in human history?

It’s not due to some recent scientific breakthrough, because we’ve had the core technology necessary to shelter-in-place (i.e., a place in which to shelter) for thousands of years. Nor are mass lockdowns a newly discovered or proven epidemiological intervention — there wasn’t any more scientific evidence of their safety and effectiveness in 2020 than there was in 1918.

While recent Internet technologies allow some people to weather lockdowns more easily than they could have in the past, every country has large swaths of the population who can’t work from home or enjoy Instacart delivery, and for poorer countries that represents the vast majority. So I don’t think that explains it either.

It’s also not because COVID-19 is any worse than other major pandemics in the past 100 years. The IMHE is currently projecting 224,546 COVID-19 deaths in the US, which is certainly tragic, but as you’ll see in the chart below, still on par with other pandemics since the 1950’s according to the CDC:

Adjusted for our current population size for relevant comparison

As you’ll see, the current projected COVID-19 death toll is about the same as the H2N2 ‘Korean flu’ outbreak in 1957–58 and the H3N2 ‘Hong Kong flu’ outbreak in 1968–69, and far less than the tragic H1N1 ‘Spanish flu’ outbreak of 1918–19. While it’s possible that lockdowns themselves have reduced the death toll of COVID-19, that certainly can’t be assumed, and there’s early statistical evidence that they have no effect on overall mortality at all.

My parents lived through these flu outbreaks in the 1950’s and 1960’s, and report to me that little attention was paid to them at the time. My mom says that all she remembers from the 1968–69 flu season was shopping for bellbottoms and peace symbols.

My grandparents weren’t yet born during the 1918 ‘Spanish Flu’ pandemic, but from what I’ve read, mass societal lockdowns weren’t even considered then either, despite the devastating toll of that virus (incidentally, H1N1 still lives with us today as a common strain of the seasonal flu). It does appear, however, that face masks were sometimes worn.

Clerks in New York working with masks on during the 1918 flu pandemic. Credit…National Archives

Societal lockdowns were never seriously considered during these pandemics, much less implemented on a global scale. Yet even today, knowing that COVID-19 carries comparable lethality to recent prior pandemics, lockdowns persist in most of the world, including countries with little ability to weather them economically.

So why did the world wait until 2020 to conduct the great lockdown experiment, and not 1918, 1957, or 1968?

In his post, Tim suggests a few theories on why American politics are so divisive today, and they all come down to structural changes in information flow:

Geographic bubbles — Because Americans are far less likely to live in politically diverse areas than they used to, a local echo chamber is created.

Mainstream media bubbles — Mainstream media has abandoned accurate and unbiased reporting as core values, and data algorithms acerbate confirmation bias in our media consumption.

Social media bubbles — We further divide ourselves into echoed information and opinion ‘tribes’ with this pervasive new technology.

I believe these same structural changes to information flows that explain our unprecedented political division also explain our unprecedented reaction to COVID-19.

Geographic bubbles — I’ve been traveling across the Western US for the past couple of months, and it’s astonishing how much lockdown and mask-wearing policies vary by state and even county. In San Francisco it’s almost universally accepted that lockdowns and mask-wearing are vital, yet in rural Utah it’s almost universally accepted that they’re silly. I don’t think that people are any smarter in one place than the other, they’re just exposed to much different information, often tracking with their political affiliation (liberals seem much more concerned about the virus than conservatives). In prior decades, when cities were much more politically and cognitively diverse, extreme attitudes towards public policy were much less likely to take root.

Mainstream media bubbles — Before the Internet started to become pervasive in the 1990’s, most Americans got their information from a small handful of media outlets that prided themselves on objective, accurate, and rational reporting. Thus, reporting on previous pandemics was much less alarmist at the time, and citizens were much less alarmed. It’s safe to say that media reporting on COVID-19 caused a worldwide panic in March of this year, which is visible in the sharp decline of public equities that month:

https://www.macrotrends.net/1319/dow-jones-100-year-historical-chart

Yet no such panic was visible in the market in the flu season of 1968–69…

https://www.macrotrends.net/1319/dow-jones-100-year-historical-chart

…nor the flu season of 1957–58 (though there was a moderate recession caused by other factors):

https://www.macrotrends.net/1319/dow-jones-100-year-historical-chart

Social media bubbles — Whatever panic and misinformation the mainstream media provided was quickly amplified and reverberated by social media, which is also fueled by sophisticated algorithms optimized for user engagement rather than objective, accurate, and rational reporting. People were further divided into extreme camps around COVID-19, with some confident it’s a hoax and others equally confident it’s a world-ending pandemic. Balanced and realistic opinions of this virus were just as unlikely to be formed in the social media environment as balanced and realistic political opinions.

In each pandemic, politicians had to deal with the potential for overrun hospitals and the terrible reality of some of their constituents dying from infectious disease, but only in 2020 did they also have to deal with mass fear caused by geographic, mainstream media, and social media information bubbles. While some critics believe that mass lockdowns were an attempt by governments to exercise their power for its own sake, I think they’re an inevitable and even necessary response to widespread public fear.

These are my theories on why lockdowns didn’t happen in 1918, 1957, or 1968.

Techie and investor. Founder at Rebel Fund and previously Pioneer Fund. Chairman of Infosurv/Intengo and CrowdMed (YC W13). Former Bain consultant. Data nerd.