The Curious Case of Japan In The COVID Times: Where It All Went Wrong For The Abe Administration
This article answers two seemingly paradoxical questions.
First, how did Japan control COVID-19 for so long while defying WHO protocols? Second, given that success, why have they recently lost control of the virus?
On the 25th of March, Japan announced that the Olympics would be postponed till 2021. The next day, 40 new cases of COVID-19 were found in Tokyo, a record increase in infections.
For many people, this confirmed their worst suspicions about the Japanese government’s response to the coronavirus epidemic. For months, the country’s experience with COVID-19 had defied expectation. Japan, similar to South Korea, Hong Kong, Singapore and Taiwan, had experienced a far slower spread of coronavirus than its western counterparts.
Yet where most east Asian nations stilled the spread of coronavirus with various combinations of early border closures, intense testing regimes, aggressive contact tracing and significant social distancing, the response of the Japanese government has been to do… not much.
For instance, the Japanese government has, against WHO instructions, done some of the least tests per capita in the developed world:
Similarly, Japan has put only minimal effort into social distancing. While the country was ahead of the curve in closing schools and cancelling public events on March 2nd — shops and restaurants remain open, while working from home is rare. As one Japanese journalist puts it:
The contrast between Beijing and Tokyo is striking and unflattering to Japan. Downtown Beijing is deserted as people try to avoid spreading the virus by staying home. Tokyo looks like business as usual, with trains and subways still packed.
The messages have been mixed. Media remains optimistic, reporting that the virus is mild and will likely taper off as summer arrives. Only healthcare workers are voicing real concern, while the public struggles to judge how to protect itself from this ‘very mild’ yet deadly virus.
Conspiracy Alone?
So how is it that Japan’s case count is so much lower than other countries, given its minimalist approach to preventative measures?
The most common theory is that Japan’s unusually low levels of testing reflect a deliberate strategy to suppress infection numbers, motivated by a desire to host the olympics in 2020. There is some evidence for this thesis. Japan’s number of deaths due to COVID-19 is suspiciously out of alignment with its number of cases, usually a sign that only the most severely afflicted are being tested, resulting in undercounting.
Australia, for instance, by the 29th of March, had 3,809 confirmed cases of COVID-19, and 14 deaths. Japan meanwhile, had only 1,693 confirmed cases of coronavirus, but a worrying 52 deaths, more than twice as many as Australia, even though they had half the number of cases. Similarly, by the time South Korea had the same number of deaths as Japan, they had three times as many confirmed cases. In fact, Japan’s death to case ratio looks quite a lot like Italy’s did at one point, which isn’t the most comforting of comparisons.
Yet it seems highly unlikely that under-testing alone would explain Japans’ low COVID-19 case count. Japan was one of the first countries to import COVID-19, hitting more than 100 cases mid February, before Italy, and well before the United States. Since then, both of those countries have seen their case numbers explode, Italy with their health system overwhelmed, and the US now leading the world in infections. Japan, in contrast, has seen only a gentle, if persistent, increase in cases.
Flushing Out Culture
If under testing alone cannot explain the divergent outcomes between Japan and other countries with slow and small responses to COVID-19, then what can?
The most common set of explanations are to do with culture. From Reddit threads to news articles, Japan is portrayed as a place where people are just naturally more hygienic than those in the West. A whole set of cultural traits are often lumped together. Japanese people wash their hands more, clean their shoes, don’t spit, wear masks, don’t litter, obey government instructions, heck, even their own health minister said their policies were successful because Japanese people were “sheep”.
Ideally, we would be able to identify which of these behaviours are more than mere stereotypes, and actually reliably slow the spread of COVID-19. There is reason to be skeptical of many of the specific claims in the “Japanese people are clean and obedient” narrative. As noted earlier the government’s mixed messages mean that even if the population is particularly obedient, they don’t know what to obey. Furthermore it’s not clear that some generalised sense of cleanliness is useful in fighting COVID-19. The virus doesn’t spread through litter or shoes. Even the extent that cleanliness does prevent disease from spreading, it’s not obvious that the culture in Japan is actually more ‘hygienic’, rather than simply being better at policing its slobs in public spaces.
Realistically, the most relevant form of hygiene in the fight against COVID-19 is handwashing, with soap, for a long time. Some do claim that Japan’s culture of cleanliness extends to the lavatory. Vox for instance, cites a survey on handwashing which claims that only 15% of Japanese people don’t wash their hands after using the toilet, compared to 40% of Americans.
Unfortunately, survey data on handwashing is famously inaccurate, with upwards of 50% more people claiming to wash their hands than actually do. Handwashing surveys are better as a proxy to national honesty than hygiene, and you can bet that if people were going around claiming that COVID-19 spread in China because people there are unhygienic and don’t wash their hands, Vox would be making the exact same criticisms.
The truth is that when it comes to handwashing with soap, most of the world is pretty disgusting, and Japan is no exception. The specific numbers vary a lot between observational studies, but almost all of them put the number of people who wash their hands with soap after going to the toilet below 60%, no matter what country you pick. This WHO study for instance, suggests that only 22% of people in Japan wash their hands with soap (compared to 52% in the states by the way). Which might explain the weird number of google searches for people complaining about a lack of soap in Japanese bathrooms.
Even when people do wash their hands, they don’t do it well, with “as many as 80% of individuals retain some disease-causing bacteria after washing”. So the next time you go to a business meeting/party/volleyball meeting, remember that at least one in two people’s hands are literally covered in fecal matter.
Can’t Touch This
All of which makes handshakes a particularly intriguing greeting. Offering them makes people like you, and a good handshake will make it more likely you get that dream job, which is unfortunate for millenials, because they’ve been told to “get a grip” on their weak handshakes. You also want to be careful because if you shake for too long (more than three seconds), people will start looking pretty anxious.
Yet if germs wanted to invent a human greeting by which to spread, they wouldn’t do much better than the handshake. For this reason, even before COVID-19 there was a movement to remove handshakes from a hospital setting, (even doctors aren’t hygienic, with less than 40% obeying correct handwashing protocols). They spread more germs than high fives or fist bumps, and those “good strong handshakes” where a salesperson crushes your hand while looking you dead in the eye are particularly effective at transferring their feces to your skin.
All of which is to say that Japan’s traditional greeting of bowing is far more hygienic than the handshake. That sort of change makes a big difference when you aggregate it across a whole population. Using some incredibly spurious numbers, the average person in the will shake hands 15,000 times in their life, or 15 times a month. Japan has a population of 127 million, and COVID-19 has been around for two months, so by avoiding handshakes over that time period Japan has prevented 4.2 billion possible opportunities for COVID-19 to spread — which is pretty truthy, but you get the picture. While leaders like Boris Johnson were busy catching COVID-19 by shaking hands with infected hospital patients, Japan was slowing the spread of coronavirus with the simple act of the bow.
Face Off
A more recent cultural habit, but possibly no less important is the normalisation of mask wearing.
In all sorts of circumstances…
Mask wearing has been controversial in the West, with many deriding them as unnecessary and unhelpful, encouraging people to wash their hands instead (which has pretty limited effectiveness by the way, only 53% of people washed their hands more frequently during the swine flu pandemic). As more evidence came out that masks actually were useful, they switched to saying that we should reserve them for healthcare workers, which is fair enough, but a non-issue in Japan where the wearing of surgical masks has been widespread for some time.
SlateStarCodex has already written an excellent post on why mask wearing is actually helpful — the summary of which is that masks are especially helpful at preventing outgoing droplets from ill patients from infecting others. They are also, to a lesser extent, useful in protecting healthy people, especially in crowded spaces. Crucially however, when the norm is for everybody to wear a mask, it is easy for the sick to do so as well, without the cost in terms of social stigma. While we don’t know exactly how effective masks are at preventing COVID-19, we do know that actions which don’t make a huge difference on an individual level can be very impactful when a whole population agrees to engage in them. Mask wearing is no exception.
Drops of Jupiter
Taken in tandem, mask wearing and bowing are powerful ways of slowing the spread of COVID-19.
But Japanese culture cannot just be reduced to a set of norms around eschewing physical contact during greetings and hiding the nose and mouth in public.
It is also about fitting lots of people onto trains.
As has been pointed out elsewhere, it is absolutely bizarre that Japan has not been harder hit by COVID-19 given that the WHO guidelines for social distancing call for four square metres of space between people…
Yet trains in Tokyo get so crowded during peak hour that people have to be stuffed in by officials.
How is it that mask use and hand shake avoidance are able to compensate for flaunting the advice on social distancing?
Religion is Good for Spreading More Than Just Ideology
Travel across the Sea of Japan, and you’ll see hints of the answer.
South Korea is now widely lauded for its response to COVID-19, but only a few weeks ago it was the source of the world’s biggest outbreak outside of China and Italy. Indeed, the fact that the virus spread so far in South Korea was something of a surprise. Having experienced both SARS and MERS the country was, like much of east Asia, relatively prepared for COVID-19.
Up till February 17th, it looked like South Korea had the virus under wraps. They had only 31 cases of COVID-19 and 30 of them had been successfully traced and quarantined.
The 31st case however, escaped the net. On the 18th of February Patient 31 attended a church meeting of near 10,000 members, even though they were experiencing flu like symptoms. By 22nd February, 1,261 attendees of the church reported symptoms of COVID-19. Within three weeks, South Korea went from 31 cases to more than 7000.
To the Japanese government, this story is typical of how COVID-19 spreads. On their government websites they claim that while on average COVID-19 spreads to two people from each person who is infected, 80% of infections do not infect others. This means that unlike seasonal flu, which tends to spread in a pretty even way, COVID-19 transmits through large clusters, created by ‘superspreaders’.
From the Japanese government’s perspective, this 80% has huge implications for how best to fight COVID-19. If only a small number of people are responsible for the vast majority of cases, then we can fight coronavirus by targeting the 20% of people who do infect others, while the lives of ordinary citizens can continue with minimal disruption. By figuring out why what differentiates superspreaders from the average citizen, the government could avoid expensive policies like broad scale testing and social distancing.
And as it turns out, superspreaders are not defined by biological differences between themselves and other people. Rather, it is to do with location and behaviour. COVID-19 is much more likely to spread in some places than others. From the same government website:
The locations where mass infections were confirmed so far are places where the following three conditions were met simultaneously: (1) closed space with poor ventilation, (2) crowded with many people and (3) conversations and vocalization in close proximity (within arm’s reach of one another).
It’s the third of these criteria which explains why the densely packed and poorly ventilated trains of Tokyo do not lead to infection outbreaks. Nobody talks on the train — not even on their phone — and of course, everybody wears a mask. Conversely, a mega-church where people shake hands, hug, and talk in close proximity is a dream infection vector for COVID-19.
Do Viruses Dream?
Being in a confined environment has caused secondary infections on the order of 10 people. The closed environment has an 18.7-fold (95% CI: 6.0, 57.9) risk.
This infection model also points us to the third reason Japan has been so successful at slowing COVID-19. While Japan has still not reached the levels of social distancing seen in many other nations, they did suspend both large scale gatherings and schools as March began — well before many places in the West . While these suspensions may not seem particularly significant relative to the lockdowns now occurring around the world — there is some evidence that large scale gatherings are an important vector for the spread of COVID-19 in the early stages of an outbreak. Aside from South Korea’s experience with patient 31, the acceleration of outbreaks in Europe (a champions league game), Iran (prayer at Qom), and India (one man leads to 40,000 being quarantined) have all been linked to large gatherings.
Furthermore, because of the exponential nature of COVID-19’s spread, social distancing measures taken early have an outside effect on the speed with which the disease spreads.
Even though Japan did not implement particularly intense social distancing, it was able to compensate by being speedy and strategic with the restrictions it did put in place.
Success Sows Seeds of Failure
So Japan implemented modest but efficient social distancing, normalised the wearing of face masks, and eschewed the voluntary exchange of germs via handshake. This explains why, for weeks, the country has seen such a gradual increase in cases relative to countries such as Italy and the US.
What it fails to explain however, is why Japan has seen such a surge in cases over the past week. If all of these national traits were slowing the spread before, why are we suddenly seeing exponential growth in Tokyo?
The answer lies in the unusual path Japan has taken to fight the virus. The Japanese government decided very early on in the fight against COVID-19 that total suppression was not a viable option, and instead decided to opt for mitigation. Essentially, it accepted that people would catch the virus, and wanted to flatten the curve. It’s unclear exactly why Japan chose this path, but it is better placed than other countries around the world to do so. Japan has the highest number of hospital beds per capita than anywhere in the world, by a long shot, and is excellent at treating pneumonia, which is the main way COVID-19 kills.
In order for this strategy to be successful, the government decided that conserving medical resources was key. The main policy implication being that the government would place limits on testing, and play down the risk of the virus. Partly, this was to ensure that not everybody with flu symptoms would rush to the hospital, wasting both tests and precious protective equipment for medical staff. In some ways, the decision makes sense. Hospitals in New York and Australia have faced just this problem, backlogs of patients with minor illnesses seeking unnecessary tests for COVID-19. As an added benefit, keeping people at home instead of encouraging them to head to hospitals prevents the disease from making its way through the medical system. This is a valid concern given what happened to South Korea during the last outbreak of MERS where a single patient infected 28 people across various hospital visits.
More strangely, the Japanese government has also been unwilling to do widespread testing throughout the community. Initially, this was due to a lack of tests, but Japan now has far more capacity than it uses. The government’s claim is that it wants to avoid an influx of false positives. Essentially, it seems to think that while tests are broadly reliable, even the small likelihood of incorrect results leads to lots of false positives when you begin testing large sections of the population. This, in turn wastes medical resources on unnecessary contact tracing, and creates panic.
Why this is a concern for just Japan, and not South Korea, Australia, or the WHO is unclear. Regardless, instead of following the WHO’s exhortation to “test test, test”, Japan has adopted a strategy of controlling infection clusters. .
Essentially, the Japanese government seems to believe that it doesn’t have to trace or prevent literally every infection. Rather, they explain in their policy documents that:
“The important thing is to minimize the spread of infection in the country by preventing one cluster of patients from creating another cluster”
To balance the spread of infection clusters against the desire to conserve medical resources, the policy of the Japanese government has been that only those presenting with very severe symptoms of infection, and direct links to known cases, are tested. If someone does turn up positive, the government has plenty of resources to engage in aggressive contact tracing and localised shutdowns.
This is the strategy that was successfully used in Hokkaido, where the Sapporo Snow Festival attracted more than two million attendees, and became the source of the largest outbreak in the country (is this a familiar story yet?). A state of emergency was declared, schools were closed, infections were located and quarantined, and within three weeks the outbreak was controlled.
Everyone’s Got A Plan Till They’re Diagnosed With Pneumonia
The Japanese government took this as a sign that their cluster control focused method was likely to be successful. They stuck with it, and Japan’s case count stayed low relative to the west.
So why then, have things suddenly gone so wrong? Tokyo has seen its new cases skyrocket over the past few days, each day beating the previous record for new infections. Some estimates suggest that Japan is undercounting it’s infection numbers by a factor of five. A government panel said that infections were “rampant”, an emergency task force has been convened, and the odds on Tokyo entering lockdown increase every day.
The fact is that even as the outbreak in Hokkaido was being suppressed, it was obvious that Japan’s strategy was based on spurious statistics, deeply flawed, and ultimately unsustainable.
Let’s first look critically at the theory underlying Japan’s response. Many of the policy documents I came across while researching this piece cited the “fact” that 80% of people infected with COVID-19 do not spread the disease to others . Remember, this statistic undergirded their whole response, because if the disease mostly transmits through superspreaders, then a few people slipping through the cracks didn’t matter, because they probably won’t spread it at all.
Unfortunately, the statistic is deeply implausible. The one paper I found investigating the variance of COVID-19’s R0 across people suggested that while superspreaders can exist for COVID-19, they are not the main transmission vector. It is also quite strange that none of the many, many papers investigating the R0 of COVID-19 noticed the quite important fact that 80% infected people fail to spread the disease. Similarly, no other government or health body has made this statistic part of their health policy, just Japan. Possibly because, frankly, it simply defies common sense. There is fairly broad consensus that on average, each person with COVID-19 infects about 2.3 others. If 80% of people don’t transmit the disease to anyone, then the remaining 20% must be infecting more than eleven people each. That seems especially ludicrous given that there is solid data from Wuhan indicating that the most common location for transmission of COVID-19 is the home. Either those 20% of people absolutely love dinner parties and large polygamous families, or there is something wrong with Japan’s statistics.
Even if the superspreader theory was completely correct however, Japan’s cluster control strategy was equally flawed in implementation. Essentially, because the system is reliant on activating when somebody tests positive in a hospital, it has difficulty locating outbreaks amongst people who do not end up in ICU, which, as it happens, represents a pretty significant chunk of COVID-19 cases. Up to 50% of all people diagnosed with COVID-19 are asymptomatic, while young people in particular often experience extremely mild symptoms. This means that outbreaks can rip through young populations before showing up in a hospital, by which point it is too late for small scale interventions.
The systematic under testing also had the effect of lulling the Japanese population into a false sense of security. Because confirmed cases grew so slowly for so long, social distancing measures were being relaxed or ignored, even as the virus gained momentum and spread through the population. On the weekend of the 24th of march, thousands attended events like the Hanami cherry blossom festival or martial arts shows in Saitama. Just days later, cases started to rise in Tokyo. Sound familiar?
In The Harshest Soil The Virus Blooms Brightest
The early implementation of moderate social distancing — combined with a culture of mask wearing and a strong sense of personal space around the hands slowed the spread of COVID-19 in Japan. Unfortunately, those same policies and cultures also allowed the government to get away with a pandemic strategy that was based on bad data, poor logic, and worse implementation. Under-testing, a deliberate part of that strategy, prevented the government from being accountable, meaning that just as COVID-19 was taking root in the country, people began abandoning the very social distancing which protected them in the first place. One hopes that Japan has managed to retain plenty of medical supplies with their conservative approach to fighting COVID-19, because soon, they might need them.
