This Daily Update is coming to you from Taipei, which normally would not be a remarkable bit of news but for the fact that I was supposed to be traveling this week; I canceled my trip.
On to the update:
From Mike Isaac at the New York Times:
The biggest story of the moment is the rapid spread of the coronavirus, which can cause a potentially fatal flulike illness and originated in China. The virus — and the global attempts to contain it — has rocked the business world, sending financial markets into correction territory, down more than 10 percent…
To try to keep their employees safe, some companies are taking preventive measures. Organizers called off MWC Barcelona, the annual global telecommunications trade show in Spain. But the giant RSA security conference still went on in San Francisco this past week, even though big companies like IBM, AT&T and Verizon pulled out of it.
Facebook canceled one of its advertising events, which is attended largely by employees, and its annual F8 conference — one of the company’s most anticipated events. That’s where it showcases its products and plans for the future to software developers. Mark Zuckerberg, the chief executive, is a mainstay at the event, giving a regular state-of-the-industry keynote address.
Karen Weise, my colleague in Seattle, had this tidbit to share: Employees at Amazon’s worldwide operations — the company’s largest division, which runs the technology and operations for warehouses, deliveries, Prime membership and physical stores, among other things — were told that they should not travel domestically or internationally “until further notice,” according to emails viewed by The New York Times.
Twitter also canceled all non-critical travel; in addition to Mobile World Congress and Facebook’s two conferences, the Game Developers Conference was postponed after the majority of major gaming companies announced they would not participate; the Google News Initiative Global Summit scheduled for April was canceled, as was Shopify’s Unite developer conference scheduled for May.
Still on the calendar:
- Google Cloud Next in April
- Google I/O in May
- Microsoft Build in May
- Apple WWDC in June
Should these be canceled as well?
Apple, China, and Data
Apple is obviously itching to get back to work making iPhones; Tim Cook said on Fox Business:
It feels to me that China is getting the coronavirus under control; you look at the numbers, they are coming down day by day by day. So I’m very optimistic there.
On the supplier side, iPhone is built everywhere in the world. We have key components coming from the United States, we have key parts that are made in China, so on and so forth. When you look at the parts that are done in China, we have re-opened factories, so the factories were able to work through the conditions to re-open. They’re re-opening, they’re also in ramp, so I think of this as the third phase of getting back to normal, and we’re in phase three of the ramp mode.
China certainly wants folks to believe that; in fact, there is already a book in the works documenting how China beat the virus thanks to the leadership of Xi Jinping. The question, though, is the same one that Bill Bishop and I discussed last Thursday: how much of Chinese data can you trust? Take this story in Bloomberg:
The pressure to get China back to work after the coronavirus shutdown is resurrecting an old temptation: doctoring data so it shows senior officials what they want to see. This phenomenon is playing out in Zhejiang province, an industrial hub on the east coast, in the form of electricity usage. At least three cities there have given local factories targets to hit for power consumption because they’re using the data to show a resurgence in production, according to people familiar with the matter. That’s prompted some businesses to run machinery even as their plants remain empty, the people said.
On Saturday, a newspaper in one of Zhejiang’s cities appeared to take aim at the practice. The Taizhou Daily published a front-page commentary criticizing local officials for narrowly focusing on power usage, arguing that hitting the targets won’t ensure economic growth. By Sunday, a link to the article on the paper’s website was no longer working…While it’s unclear how wide the problem of doctored data is in China, there are signs that electricity usage has become a focus for more than Zhejiang. The official Xinhua News Agency last week published a story about production resuming in Guangdong, China’s largest provincial economy, using power consumption as the main evidence to show how quickly things were getting back to normal.
Beijing has tried to assure the public that the economic restart is going smoothly, touting official statistics that measure the proportion of firms in each province that have resumed work, but don’t account for the capacity they are operating at. Data released last week by the country’s top economic planner showed industrial hubs like Shandong and Guangdong have more than 70% of companies back at work, while in Zhejiang the rate is over 90%.
From what I have heard, local officials are in a bit of a bind: on one hand, they feel the pressure to have no new infections; on the other hand, Beijing is strongly signaling that the economy needs to be restarted as soon as possible. Given these conflicting incentives, it’s not surprising that some may be fudging the data, and it seems a bit naive to think that electricity usage — and not, say, the number of new cases — would be the only data that is fudged. Then again, Cook surely wants China back to work as well; believing the official numbers continues to be the path of least resistance.
What We (Might) Know
Before I go forward, two very large caveats are in order: first, no one knows for sure what is happening in China, and second, I am not an epidemiologist; I have, though, been observing China for years, and following COVID-19 extremely closely since January, both for professional and personal reasons, and wonder if there might not be something in the intersection. To that end, the following is my best guess as to what is going on, and what the implications might be for tech companies and their conferences.
First, it seems likely that the number of cases was dramatically larger in China than was ever reported, both because the data is generally unreliable, and also because it seems likely that most people with mild cases were never tested (much less those without symptoms). Another axiom that Bishop and I discussed was the importance of observing what Chinese authorities do, not what they say, and the drastic quarantine efforts undertaken in Wuhan in particular fit this hypothesis.
Secondly, it seems likely the fatality rate is lower than what was reported in China. Consider the failed quarantine of the Diamond Princess cruise ship, which allowed for a controlled sample. According to the latest numbers from Japan, of the 3,711 passengers and crew members on board, 705 were infected; of the 705 infected, 392 were asymptomatic; six have died. That is an infection rate of 19.0% (10.6% with symptoms), a fatality rate of 0.8%, and a mortality rate of 0.2% (fatality rate is what percentage of the people who are infected die; the mortality rate is what percentage of the overall population die). These numbers — both the number of infected and especially the numbers of deaths — are likely to rise over the next few weeks, and of course anyone capable of going on a cruise is in better health than the elderly and those with underlying health problems that appear to be at greatest risk; still, the fatality rate appears much lower than in China, even as the virality appears greater.
This is mirrored in South Korea, which is experiencing the largest outbreak outside of China — at least the largest outbreak that we know about. This is an important caveat both because some countries may be worse (particularly Iran), but also because the South Korean numbers, thanks to the country’s excellent infrastructure, health care, and free press, are likely to be amongst the most reliable in the world. There the fatality rate is 0.4%, even lower than the Diamond Princess. Again, these fatality numbers are going to get worse, as it takes multiple weeks for COVID-19 to become fatal, but it is hard to imagine a scenario where they are as bad as it once seemed in China.
In short, it seems likely that SARS-CoV-2 (the virus) is more viral than first thought, while COVID-19 (the disease caused by the virus) is less fatal than first thought. This makes sense intuitively — the more people that have the disease means a larger denominator when it comes to calculating fatality rates — and it appears to be reflected in our best available data sources.
This is both good news and bad news. Start with the latter: an outbreak in the rest of the world, including the United States, is very likely impossible to stop at this point. There was one opportunity to stop the virus, and it passed some time late last year; the takeaway from China’s massive quarantine effort should probably be that it was ineffective. Sure, China says that Hubei province was uniquely bad, but it seems hard to believe that the rest of the world would start seeing numbers worse than the rest of China. Again, this is where it is critical to observe China’s actions, and not their data: the country is very much acting as if it has given up on detainment, at least at the national level, and restarting the economy appears to be the top priority — as it should be if, in fact, shutting down entire cities didn’t work.
The good news is the inverse of that — and to be very clear, the “good” is relative. First, as I noted, it seems likely that the fatality rate is lower than once feared, and second, economies probably shouldn’t shut down in response to outbreaks; this will relieve suffering of another sort.
All that noted, there is one more number that is harder to get ahold of, but matters quite a bit: it does appear that a significant percentage of those who are affected by COVID-19 suffer from very serious symptoms that require hospital-based medical care, even if they don’t die (this is decidedly not just the flu). That means that while a large proportion of the population may get COVID-19 eventually, it does make a real difference as to how quickly that happens if we are to ensure that hospitals aren’t overwhelmed. In that light, the rapidity with which the U.S. closed down flights from China was a real win: it almost certainly slowed the speed with which the virus entered the U.S., even if it didn’t ultimately stop it.
In my opinion this should inform how companies think about large gatherings like developer conferences. On one hand, “Keep Calm and Carry On” is probably the order of the day from a big picture perspective; that augurs for keeping on with travel and conferences and whatnot. At the same time, said gatherings, at least for the next couple of months, seem like an unnecessary risk of not necessarily spreading the virus (which is probably fairly inevitable) but of accelerating the spread in a way that taxes healthcare systems and communities. I would cancel them (a similar rationle is why I canceled my trip).
That said, please note the caveats above. We are not only dealing with imperfect data, I am an imperfect narrator. Moreover, there is still time, particularly for conferences in May and June; that said, it may take that long to truly understand what is happening on the ground, as the next few weeks and months will likely get worse before they get better.
I recommend this post from Elad Gil that he wrote for startups; it contains the latest overview of COVID-19, what to expect going forward, and best practices for companies. Gil has a full citation list for further reading, and he has committed to keeping this Google Doc updated.
This Daily Update will be available as a podcast later today. To receive it in your podcast player, visit Stratechery.
The Daily Update is intended for a single recipient, but occasional forwarding is totally fine! If you would like to order multiple subscriptions for your team with a group discount (minimum 5), please contact me directly.
Thanks for being a supporter, and have a great day!