Explaining Japan’s State of Emergency

With Prime Minister Abe’s long-awaited emergency declaration for Tokyo, Kanagawa, Saitama, Chiba, Osaka, Hyogo and Fukuoka yesterday, the response to the COVID-19 pandemic in Japan has entered a new phase. As many people have pointed out, these new measures are not a “lockdown” of the sort that has been implemented in many countries around the world, with legally mandated closures and stay-at-home requirements which carry actual penalties for non-compliance. The Japanese legal system currently makes it almost impossible to implement similarly stringent directives (because the entire nation was compelled to extremely senseless self-sacrifice before and during the Second World War). Instead, for the coming month, the government is requesting voluntary compliance with more extensive social distancing measures than requested previously. Services deemed “essential” like supermarkets, convenience stores, clinics and hospitals, hotels, as well as manufacturing operations, are requested to remain open. As previously, people are asked to avoid establishments where large numbers of people gather, such as bars, clubs, live music venues, and karaoke, and if they have to eat out, to avoid doing so in large groups. Oddly, perhaps, public baths are asked to remain open. Many public venues such as movie theaters, museums, libraries, and department stores have already closed temporarily in Tokyo in response to earlier requests for distancing, and may be asked to do so again. But nobody will be forced to do anything and anyone can refuse to comply, and at present the order only applies to seven out of 47 Japanese prefectures. Alarmingly, there are already numerous reports of people fleeing Tokyo for other prefectures in order to evade the emergency restrictions. Partial shutdowns accompanied by ineffective monitoring of travelers made things immeasurably worse in China, Italy, and elsewhere, as infected people seeded new clusters in their home countries and abroad.

The seven prefectures and metropolitan areas included in the new emergency declaration. Source: NHK

Prime Minister Abe is quoted as saying, “If all of us make an effort and try to reduce person-to-person contact by 70 percent or, preferably, 80 percent, the increase in infections will peak in two weeks and shift to a decline.” There is a lot of uncertainty surrounding this, especially voluntary compliance, which so far has been spotty at best in Japan. We wish stronger and more binding measures, as well as more forceful messaging, had been put in place weeks ago.

Agence France Presse editor Richard Carter tweeted:

On the one hand, after the disasters of March 2011 we saw that Japanese people will readily cooperate with inconvenient emergency measures when the need has been made clear to them. In 2011 this primarily took the form of widespread voluntary energy conservation (setsuden) nationwide — shops turning off half of their lights, illuminated advertisements being turned off in train stations and building exteriors, escalators being shut down, as well as households limiting their own power consumption — which led to about a 20% decrease in overall electricity consumption. Voluntary restraint (jishiku) was also requested then to limit celebrations, other large social gatherings, and extravagance in general, and people complied.

This is in some ways a recent precedent for the still moderate sacrifices being requested now. A few days ago TBS released poll findings which showed that about 80% were in favor of an emergency declaration. But will they actually comply? We’ll know in about a week, though we don’t think the necessary ground preparation has been done by the government and communicated to the public. This includes providing additional free child care, subsidizing workers who have to stay home, as well as their employers, and easing bureaucratic regulations that make it hard to submit paperwork online, among others. In most companies, universities, and other institutions, most staff are still required to be physically present in order to be able to stamp physical documents with their hanko seal. In the current health crisis this can only be seen as barbaric. Some relevant support measures have been talked about and promised, but none of it is actually in place today. We contrast the apparent lack of care in Japan with the thorough planning done by the Singapore government, which last Sunday announced very strong binding measures in response to a “second wave” of infection (something also being experience in Hong Kong now). These guidelines have been communicated unambiguously using official outlets as well as the media, and we think they should be considered best practice.

Japan case data as of April 6th. Source: NHK

There has been a clear uptick in COVID-19 cases in Japan as a whole since late March, led by Tokyo. There is wide agreement that this is due to masses of people having gathered closely in parks three weeks ago for cherry blossom viewing parties and infecting each other, despite half-hearted official requests for social distancing. While some important uncertainties remain, and the central government’s communication efforts are still lousy, data availability for Japan has improved overall in recent weeks. The Tokyo Metropolitan Government’s multilingual COVID-19 web site has become a much clearer and informative official source, better than anything the national government has provided. As of today it shows 1196 positive cases in Tokyo, with a sharp upward trend since two weeks ago, exceeding 100 new cases per day over the past weekend, declining somewhat on two subsequent days, and reaching 144 today. The nation as a whole now has over 4480 cumulative cases, with over 300 new cases reported every day since April 3rd, except on April 6th, which saw 241 new cases. The very well-run independent and bilingual Japan COVID-19 Coronavirus Tracker, which has an extremely detailed prefecture-by-prefecture open database, has become our go-to source of vetted case number data for Japan.

Source: Japan COVID-19 Coronavirus Tracker

As we noted in earlier articles, many people have concluded that the lack of extensive testing in Japan has led to a serious underestimate of the actual number of COVID-19 cases and the rate of contagion here. Dr. Kenji Shibuya, director of the Institute for Population Health at King’s College, London, was quoted in the Washington Post yesterday saying, “It’s too late…Tokyo has already entered an explosive phase, and the only way to stop the collapse in health care was to lock down the city as early as possible.” He considers the sudden uptick in cases over the past week to be clear evidence that the country’s limited testing strategy has been a failure, the magnitude of which will become clearer as testing and discovery scale up.

While testing in Japan has increased, it still lags far behind every country that has succeeded in flattening their outbreak curve. The Prime Minister announced a few days ago that testing capacity would be increased from 7,500 to 20,000 per day, but it’s not clear how soon this can be achieved. As of April 7th, a total of 55,311 people have been tested in Japan, on almost all days fewer than half of the current stated capacity of 7500 tests per day. Germany, by contrast tests 500,000 per week.

In Japan as elsewhere, the inability to test everyone who has died of pneumonia-like respiratory failure also leads to a possible undercount of COVID-19 cases. Conversely, the widespread worldwide practice of categorizing every mortality in which COVID-19 is present as having been due to that virus has been criticized by some as leading to a possible overcount of COVID-19 deaths. But it’s been pointed out that people in many countries who die while suffering from normal seasonal flu are recorded as flu deaths regardless of whatever other health problems may have also contributed. It’s not a controversial practice, though the implications for case severity tracking must be kept in mind. All-in-all, we think it’s clear that there is much greater public health risk from underestimating the scale and growth of the pandemic than from overestimating it.

Japan COVID-19 case severity breakdown. Source: NHK

There is a serious concern among Japanese government and healthcare specialists that not enough intensive care unit beds will be available if the number of serious cases continues to grow at its current pace. The proportion of ICU beds per 100,000 people is much lower here than in many other developed countries (5 per 100,000 in Japan as opposed to 12 in Italy and about 30 in Germany, for example). This independent site shows the current status of ICU bed availability in each prefecture. In at least five prefectures, including Tokyo, the number of COVID-19 cases already exceeds the number of ICU beds. While according to NHK, only about 4% of COVID-19 cases seen in the country so far have become severe enough to require ICU treatment, the potential of severe cases outstripping capacity is looming.

In response, the government has reportedly obtained cooperation to provide about 10,000 hotel rooms in Tokyo and 3,000 in Kansai, as well as 800 rooms in the Olympic village in Tokyo, that can be used by patients with mild symptoms. Patients began to be moved to a hotel in Tokyo’s Chuo Ward yesterday. As we noted back on March 24, although many observers credited various Japanese behavioral norms, like mask-wearing and hand-washing, for having kept COVID-19 numbers relatively low until then, we felt that simple luck had also played a role. What happens now? Japan could continue to be lucky. The current spike could prove to be an anomaly, and the new voluntary measures might be met with widespread compliance sufficient to nip the surge in the bud. A recent study, which may have influenced the decision to announce the state of emergency now, posits a harrowing worst-case scenario that shouldn’t be ignored, however. If the voluntary measures fail to curtail more than 60% of people’s trips outside their home, the authors conclude, then the healthcare system will collapse on or around April 26th due to lack of ICU capacity, and 500,000 people will die. The authors seem hopeful that the 60% decrease can be achieved, noting unpublished research which shows that school closures since March 2 decreased contact frequencies among children by 40%, while voluntary event cancellation since February 27 decreased it among adults by 50%.

As with every study, this model makes basic assumptions that may not be fully accurate and is based on Chinese data that has inherent uncertainty. The Prime Minister has asked the nation to reduce social contact by 70 or 80 percent for the next month. But without clear incentives for companies to allow their employees to remain home without incurring financial hardship, we think these targets will be difficult to achieve. It certainly won’t be as effective as it would have been if stronger and more binding measures and more forceful messaging, like those being implemented elsewhere around the world, had been put in place in Japan weeks ago. Now we seem to be counting too much on luck.