Nice write-up!
A few thoughts re: Scott Alexander & Rob Wiblin on prediction.
Thanks for the great effort for bringing a summary that really makes me feel I don't have to continue digging for new information. If this work is going to be continued, I will just wait for your new updates, and then finish looking for local news, as things on my country are largely different from yours. Keep it up!
This is fantastic, thanks. My only wish is that it was more often, ideally weekly or 2x per week for now and then decreasing in frequency as the situation becomes more stable. This is because with it coming out less often and there being lots of seemingly significant news in the meantime, I still feel the need to stay up to date - but I think if I knew something of this quality and comprehensiveness was coming out often I wouldn't. If it was subscription only, I'm pretty sure I'd pay for it, too - though that's probably not a good thing as it'd reduce the number that can benefit.
Thank you so much for this summary!
I don't like in Rob Wiblin's post that it presumes that the <5% forecasts of serious pandemic were wrong, and his forecast of ~80% was correct. (And I'm writing it here rather than on the original post because it seems wrong to visit a FB page of a guy I've never heard about just to criticise him.)
It’s been almost a month since my last COVID Consolidated Brief and I hope you are all doing ok. I’ve personally been settling into the new normal. On the other hand, I’ve witnessed first hand some of the risks that might be coming with COVID. I had to take shelter in a tornado warning for the first time in my life. While the tornado and the destruction were luckily quite minimal, there was a power failure for about a day and social distancing made it a lot harder to wait out the power failure in a nearby library or Starbucks. Overall, I’m lucky my life is so safe that this is the biggest problem, but I am worried about people who might be a lot less likely and face strong hurricanes or wildfires while also having to maintain social distancing. More on this in a bit.
If you’re just joining us, I follow COVID-19 a lot and this is my third semi-regular installment of a public consolidated brief that tries to consolidate everything I read into one short, actionable list so other people can stay up to date without reading a ton on their own. For this issue, I spent over 25 additional hours trying to get to the bottom of everything so that you don’t have to. This way I can save time and fight research debt and save you time from having to read all of this yourself. That being said, do keep in mind that I am not an expert and I have not been able to cover everything going on - I had to be fairly selective to make this brief actually somewhat brief.
I’m not sure how often I will do these, but I still intend to do them as I am able. Maybe it will be a monthly newsletter. Maybe I’ll be able to do it every other week. We’ll see!
Previously:
See also:
Doing Your Part! How You Can Stay Safe and Help the Fight!
Masks
Masks are a good idea - tell your friends! The opinion on masks has changed a lot since I last reported about a month ago.
WHO is now onboard: “The World Health Organisation says it supports government initiatives that require or encourage the public wearing of masks, marking a major shift from previous advice amid the Covid-19 pandemic. [...] The WHO added that surgical masks should be reserved for medical professionals, while the public should use mainly cloth or home-made face coverings.”
The American CDC is now onboard: “CDC recommends wearing cloth face coverings in public settings where other social distancing measures are difficult to maintain (e.g., grocery stores and pharmacies), especially in areas of significant community-based transmission.”
A lot more information on masks is now available in Thomas Pueyo’s “Coronavirus: The Basic Dance Steps Everybody Can Follow”.
Masks may even become mandatory as a part of the reopening plans - see more below.
Giving
Want to help give money to help people most affected by COVID-19 have a chance to get back on their feet? GiveDirectly is now helping you give cash directly to those in most need.
Want to give money to helping analyze and treat COVID? I would donate to the Center for Health Security at Johns Hopkins, which researches biosecurity and has been tracking COVID since early January. You can donate here, or you can donate through the Effective Altruism Funds.
~
My previous 29 Mar Brief has a bunch of advice on how to stay safe and contribute.
A Glance at The Latest Situation
The last time I wrote, we were just about reaching the one millionth case worldwide. Now we’re above one million cases just in the US and we’ve exceeded three million cases worldwide.
Here’s the latest based on cases...
and deaths...
It seems that the UK has now likely peaked and is coming down, whereas it is still too early to tell for the US. However, the descending part of the curve seems to be much slower than the sharp ascending part of the curve. It seems increasingly likely (and is even now acknowledged by the IHME) that we won't get a steep, bell-curve-type “mirror image” decline in cases / deaths as previously hoped for.
It looks like Austria, Australia, New Zealand, and Norway are joining Taiwan, South Korea, and Hong Kong as potential case studies in successful handling of COVID. On the other hand, previous darlings Japan and Singapore now look to each be facing a moderate outbreak.
~
Indeed, in good news, New Zealand has already declared victory:
~
Though Singapore’s previous victory is now revoked and shows security may not be as absolute as it seems. They previously had ~20 cases per day in mid-March but jumped to ~1000/day in mid-April, declared a lockdown on 7 April and added additional measures on 21 April, and have seen cases drop to ~500/day.
~
Here’s another way to look at the numbers, which looks at new deaths per capita, averaged over the past week to smooth out some (but not all) reporting issues… makes you wonder what the heck is happening in Belgium and Ireland (if it is anything more than differences in how countries report death stats).
~
Reporting deaths is hard. Looking to “excess deaths”, the death toll could be even worse than currently imagined:
~
Also, Business Insider reports that latest forensics show that people died from COVID in the US earlier than we thought:
~
As a further sign of how difficult all this death-counting is - “NYC death toll jumps by 3,700 after uncounted fatalities are added”:
~
In “Coronavirus: Learning How to Dance”, Tomas Pueyo also breaks trends down on a country-by-country basis (numbers shaded based on relative size of new cases within that country):
(see bigger)
It’s worth taking a look at the bigger image, which shows that while outbreaks may be steady in more developed countries, they’re just beginning to start taking hold in less developed countries. Here, we can see a lot of countries are just starting to see their outbreaks, while other countries have peaked and declined. From here we can visualize where all the countries are on a curve from handling the virus:
(see bigger)
~
Here’s a world map of current COVID deaths per person:
~
It’s easy to get lost in the numbers, but the amount of death we’ve seen so far is a lot - the amount of deaths in the US over the past few months far exceeds American deaths in September 11th and the Iraq War and recently also just exceeded the total number of American deaths in the twenty years of the Vietnam War (source):
And COVID has become the leading cause of death in the US:
...So Just How Bad Could This All Get?
Just as we might start getting good news about coronavirus, some other disasters might make it worse. For example, Colorado State University just issued a hurricane season forecast that doesn’t look good:
Needless to say, a hurricane like Katrina, Harvey, or Maria is devastating enough on its own, but would combine with already overloaded hospital systems in a very bad way. And as if hurricanes aren’t bad enough, there will almost certainly be large-scale wildfires again as there were last year.
Also, needless to say, social distancing may be difficult during evacuations and this could lead to new outbreaks facing an even more overburdened hospital system.
~
Many conflicts could be potential time bombs amid the coronavirus. I’m still particularly worried China may take the opportunity to be more aggressive in the South China Sea, especially after China rammed a Vietnamese fishing boat at the beginning of the month. China has also been bullying Taiwan with military flybys. The US has already called for $20B in new military spending to deter China.
Meanwhile, Politico reports that the US State Department has continued to warn that Russia, China, and Iran are pushing a host of matching propaganda disinformation messages, including that the coronavirus was an American bioweapon.
Americans are increasingly negative on China. A new Pew Research survey finds 66 percent of Americans say they hold an unfavorable view of China, up 6 points from the previous year. And of course, recent ads from the Trump camp referring to Biden as “Beijing Biden” and Biden retaliating by tying Trump to China suggest this may only get worse.
While I am all for due criticism of the Chinese government and think there is a lot to rightfully criticize, I do worry that declining American sentiment toward China could risk us closer toward armed conflict.
~
To make matters worse, the World Food Programme warns of “multiple famines of biblical proportions”:
This is really bad.
~
Another question - less important but still of interest - will those famines happen in the developed world too?
On 27 March, I wrote that Metaculus had put the probability of a food shortage in a major US, UK, or EU metropolitan area before 6 June at 30%. I thought this was insanely high and offered my own prediction of 5%.
I see Metaculus has now come to their senses and now has a median prediction of 6%. However, now things might be starting to feel a bit different. It’s hard to get more blaring than Tyson Foods takes out full-page ad: 'The food supply chain is breaking'. This seems to mainly be limited to meat production in slaughterhouses, of which a lot have had to be closed or limited due to spread of COVID among slaughterhouse workers.
The Wall Street Journal reports:
NBC reports:
Still, a meat shortage doesn’t mean a food shortage writ large, and 6 June is now just a little over a month away and I expect we have more than a month of slack in food production, so overall I continue to put very low odds on this question (personally dropping now to 3%).
~
Covid-19 May Worsen the Antibiotic Resistance Crisis: A bit of a double-whammy where patients ill with COVID are prescribed antibiotics to protect against other possible infections, thus increasing antibiotics use and thus increasing potential antibiotic resistance… at the same time antibiotics manufacturers are slowing the development of new antibiotics to focus on anti-COVID treatments instead.
And How Do We Get Out of this Mess? Vaccines, Treatments, Testing, Tracing, etc.
The Grand Reopening
Getting out of this mess is top of mind as the biggest news of the moment seems to be how we handle re-opening.
The United States
NBC reports that according to US Vice President Pence sixteen US states have unveiled “formal reopening plans” to lift coronavirus restrictions.
It’s worth reiterating the “National Coronavirus Response: A Roadmap for Re-Opening” plan from the American Enterprise Institute that I last covered in my previous brief as it is now basically the official plan endorsed by Donald Trump. As I mentioned before, this proceeds in three broad phases:
Phase I: Slow the Spread. Widespread school closures, work-from-home, close malls and gyms, and limit restaurants.
Phase II: State-by-State Reopening. Individual states are able to move to Phase II as they are identified to be able to safely diagnose, treat, and isolate COVID-19 cases and contacts. Testing must be scaled up rapidly. These states can gradually reopen schools and businesses, but will likely need to maintain some degree of physical distancing and limitations on larger gatherings. Older adults will also need to remain at home.
Phase III: Lifting restrictions. Phase III will also be gradually reached on a state-by-state basis “once a vaccine has been developed, has been tested for safety and efficacy, and receives FDA emergency use authorization” OR “there are other therapeutic options that can be used for preventive or treatment indications and that have a measurable impact on disease activity and can help rescue very sick patients”.
~
The point of these lockdowns has been to buy us time to (1) build up hospital capacity to handle a larger future wave and (2) build up testing/tracing/isolation capacity to be able to more finely quarantine just those with COVID. Insofar as we’re accomplishing (1) and (2), we can start reopening the economy.
My current guess about re-opening is that there won’t be a single binary “everything back to normal” event like it seems people are conceptualizing it. People think of “reopening” as “everything goes back to how it was before the coronavirus”, but that seems quite unlikely.
Instead, I expect the country to gradually reopen, largely in reverse of the way that it closed, except a lot slower. That is, the things that were closed last (e.g., parks) will be reopened first - potentially really soon - whereas the things that we closed first (e.g., concerts, conferences) will be reopened last - and potentially not until after we have a widespread vaccine.
For a look at what the very first step might be, look to the new “Safer at Home” policy of Colorado that basically reopens private gatherings with less than ten people, one-to-one real estate home showings, curbside pickup, and not much else:
~
Some governors and mayors have been a bit more gung ho about reopening, however.
The semi-serious Georgia is permitting a reopening of “gyms, fitness centers, bowling alleys, body art studios, barbers, cosmetologists, hair designers, nail care artists, estheticians” with “screening workers for fever and respiratory illness, enhancing workplace sanitation, wearing masks & gloves if appropriate, separating workspaces by six feet, teleworking if possible & implementing staggered shifts”.
Even Trump seems to think this is too much, saying “I disagree strongly with [Governor Kemp’s] decision to open certain facilities which are in violation of the Phase One guidelines [...] I think spas, and beauty salons, and tattoo parlors, and barber shops, in Phase One … is just too soon. … They can wait a little bit longer.”
~
Tennessee is also reopening restaurants and some other businesses:
~
It gets worse - the widely mocked Las Vegas Mayor Carolyn Goodman just wants to offer up Vegas as a "control group" to measure the effects of lifting restrictions.
~
Maryland has a proposal that recommends lifting restrictions only after COVID deaths and new hospitalizations have seen a consistent two-week decline - more conservative than Trump’s guidance of two weeks of decline in new cases.
~
Threading the needle between Georgia and Maryland, Texas is going with a 25% plan:
~
The Bay Area is extending their lockdown through the end of May, but now allowing drive-in religious services (stay in the car), one-on-one residential real estate viewings, golf courses, and driving ranges.
~
So when all is said and done, what might things begin to look like? CNN suggests that "America's 'new normal' will be anything but ordinary", potentially seeing some of the following changes:
Another take from the Washington Post asks “How much of our lives will coronavirus change?” and speculates:
~
It's hard to know when and how we should be reopening, and I imagine that many governors are willing to take on some additional death and some risk of disaster in order to reopen their states.
Furthermore, it’s also unclear how much people will voluntarily take up these recently reopened businesses versus stay at home. While Georgia is allowing stores to reopen, one reporter reached out to twenty different small businesses and not one of them said they had plans to reopen this week.
...All of this could introduce a lot more uncertainty into the future of the effects of the coronavirus.
Again, caution is warranted about making predictions about how things will unfold:
~
We also have to wonder about the political will to roll this out. So far, we’ve done reasonably well at adhering to social distancing for long periods of time with only minimal protesting. I hope this can last for many more months, but I’m unsure. As Ezra Klein summarizes - "I’ve read the plans to reopen the economy. They’re scary. There is no plan to return to normal.":
~
How dangerous will coming out of lockdowns too early be? The real point is that we just don’t really know, as we don’t know how much voluntary social distancing will remain in place, among many other uncertainties. But previously we knew that it actually took an exceptionally strong lockdown to keep deaths down, so reversing that back to a not-all-that-strong lockdown could be disproportionately bad. Hence “New Model Shows How Deadly Lifting Georgia’s Lockdown May Be”:
One potential cautionary tale might be Japan’s northern island of Hokkaido:
However, again, uncertainty remains a concern. If we make dire predictions that overestimate the return to normal activity or underestimate potential beneficial effects of mask wearing, weather, etc., it’s possible things might not be as dire as they sound, which could lead to people distrusting these expert predictions more. ...On the other hand, things could end up worse than projected.
Colleges
For another question, when and how will Harvard reopen for the fall term? Here’s how they’re thinking about it:
Here’s an alphabetical list of colleges that have either disclosed their plans, mentioned them in news reports, or set a deadline for deciding. It looks like most of them are aiming to have in-person classes.
...I’m personally wary about how a large campus is supposed to reopen. After all, the risk of spreading COVID is proportional to the amount of people, their density, and the length of contact and college has a lot of all of that.
Indeed, one study found that “[t]he average student can “reach” only about 4% of other students by virtue of sharing a course together, but 87% of students can reach each other in two steps, via a shared classmate. By three steps, it’s 98%.”
Even on campus, presumably students will have to be aggressively pre-tested and quarantined before joining the campus population. Also, at least some steps to ensure isolation and some social distancing will have to take place on campus even after all of that.
Facebook
Another place to look to might be Facebook. Per their latest announcement, they will be (1) requiring nearly universal work from home through at least the end of May, (2) cancelling all business travel through at least the end of June, and (3) and cancelling all large physical events with more than 50 people through June 2021.
The United Kingdom
Buzzfeed News reports on the UK’s reopening plan:
South Korea
South Korea has remained relatively open, albeit with substantial restrictions in place.
One contact says people don't really notice the impact of COVID any more, though February and March felt pretty chaotic. Working from home is encouraged, esp. for employees that would have to use public transport, but many people come to offices. The next few weeks will be critical: will see if clubs and grocery stores can reopen in Seoul without causing a huge spike in infections.
COVID-19: Testing, Isolation, Geolocation in Korea gives a bit of a glimpse at how seriously South Korea is taking this:
Africa
It looks like some African countries are coming out of lockdown:
However, lockdowns in developing countries often force people to choose between starvation or disease.
Worse yet - for many African countries, this is still just the beginning:
Vaccines
What kind of treatments and vaccines might we expect? Derek Lowe outlines “The Order of the Battle” where first we try repurposing existing drugs (e.g., remdesivir, hydroxychloroquine, azithromycin, falapirivir, ivermectin), then monoclonal antibodies, then vaccines, and then potentially new treatments. It’s possible existing drugs or new treatments could greatly reduce the danger of COVID and allow for faster reopening, but it would take a vaccine to truly make it go away.
So let’s zoom in a bit on vaccines. Keep in mind that producing a vaccine in 1.5 years is by far the landspeed record for making vaccines. The ebola vaccine took five years and as I researched in 2017, it historically takes on average thirty (!!) years. We expect unprecedented acceleration here, however, because we have unprecedented resource mobilization and an unprecedented willingness to not undergo all the same rigorous safety measures.
Derek Lowe provides updates on “Coronavirus Vaccine Prospects” - we need to make a vaccine (of which there are several possible types to try), it needs to be effective at treating COVID, it needs to be safe (not introduce any new diseases or symptoms), and it needs to be effectively manufactured and rolled out at scale. Any of these three prongs could slow down vaccine progress.
And Derek Lowe also provides “A Close Look at the Frontrunning Coronavirus Vaccines As of April 28”, tracking eight major efforts:
~
The aptly named VisualCapitalist tracks the status of some ongoing treatments and vaccines:
~
A much more detailed tracker of over 280 treatments and vaccines is available from the Milken Institute.
Gaze into the Crystal - The Latest Modeling and Forecasting
Metaculus has been doing some cool stuff, and I always like checking on their latest dashboard:
It looks like compared to the numbers from 2 April, the lower 25% and median bounds of estimated infections and cases confirmed by testing have gone up. But some good news - the median estimated number of deaths has gone down slightly.
~
Another prediction dashboard, from Good Judgement Inc., also looks at cases and deaths, though over a slightly different timeframe, and seems to come up with fairly similar results:
~
Looking at the actual models themselves used by the expert forecasts above, we can see they actually struggle a bit to agree:
Calibration of these models also continues to remain a concern:
~
More good news - the timeline for the vaccine has gotten much more optimistic over the past month, with the lower 25% and median both shifting six months earlier, the upper 75% shifting 1.5 years earlier, and the chance of a vaccine waiting until after March 2027 fell from 8% to 6%.
On the other hand, the Good Judgement forecasters appear a lot more pessimistic, putting almost double the odds (50% vs 25%) on a vaccine not appearing by April 2022. (Note that the scale of the distribution is different, but I’m doubtful this matters much for the timeline at this scale.)
~
Metaculus has also been trying to forecast the easing of lockdowns. As seems confirmed by the reporting above, they’re imminent. However, be careful to not interpret easing of lockdowns as anything but the first step in a gradual process - the way these questions are worded, any easement will count as a positive case, even if it is a lot less than would make it possible to enjoy the world to the extent we used to.
To get a bit more granular, Metaculus currently has 47% odds on “most of the classes for courses at Harvard College that would usually be scheduled to occur on September 2nd have in person instruction on September 2nd, 2020.”
Some other questions in a similar vain we might want to ask to get at the granularity of reopening… and my personal predictions:
I made these about Chicago because that’s the area where I live and I know well and because Chicago and Illinois have historically been more cautious about COVID. I added these questions to Metaculus to see what others think.
~
Laypeople can make a prediction too… it looks like most people are expecting a full “return to normal” sometime this year. I personally find it unlikely we would be able to see a “return to normal” until after we have a vaccine, which does not seem likely to happen in 2020 at all.
~
Even cooler, Metaculus will be taking these forecasts head-to-head with experts:
~
If you want cool and useful plots about COVID cases, deaths, and testing broken down by US state, county, and city, check out covid19watcher.com!
Now Just What are the Tech Overlords up to?
Google and Apple are joining forces on contact tracing. This is big news, as Google’s Google's Android and Apple's iOS have approximately 100% of the mobile OS market share with 3 billion combined users. The plan is for Apple and Google to release interfaces next month for health officials to build apps and for contract tracing to be enabled over the next few months.
The app works by running Bluetooth in the background and broadcasting beacons that are logged by nearby phones. Once someone is diagnosed with COVID, they can consent to sharing the past two weeks of beacon logs with health professionals, who can then broadcast anonymous alerts to those who have been in contact.
Google and Apple emphasize that the system is built on a decentralized stack that does not broadcast any data without user’s permission and does not use location data at all (only data of who you came in contact with, not where).
However, the app needs about 60% of the adult population to use it for it to be fully effective at controlling COVID on its own, but it could still be an important part of a larger contract tracing effort.
~
Bill Gates is spending billions of dollars building manufacturing facilities for many vaccine candidates, even though many of these facilities will go unused:
~
People sometimes criticize billionaires for not donating enough. Jack Dorsey gets it and is donating a full billion (~28% of his wealth) to fund global COVID-19 relief. He’s tracking his donations here and it looks like he’s already dispersed $8M.
Now Let’s Talk Policy Response
Polls consistently show that Americans still strongly support social distancing:
In fact, support among freedom-loving Americans seems to largely be in line with strong support in other countries:
~
How can we scale up vaccine progress? There are many ways but one compelling way might be option-based guarantees:
A Bit About Life Under Quarantine
COVID certainly is taking a toll on all of us. Pete Davidson and Adam Sandler have teamed up to sing about it. Self-reported well-being in the US is at a 12 year low:
~
What does all of the current modeling imply for your summer vacation?:
Of course it depends on the month - travel in August seems a lot more likely than travel in June… and I don’t expect you’ll be going anywhere internationally.
~
Baseball might be coming back this year:
~
The real important question - how long until we run out of new TV? Luckily it looks like Netflix can keep producing shows well into 2021.
~
The latest victim of COVID-19? The hotel mini-bar.
If You Still Own Envelopes, Check Their Backs - Here’s the Latest Cost-Benefit Analysis
A new paper has come out that may be the first proper attempt to evaluate COVID policy in terms of wellbeing and suggests that net benefits of releasing lockdown will be positive from June:
The paper has some shortcomings - among others, it would be good to have more probabilistic analysis given the highly uncertain inputs, the effect on mental illness remains guesswork, the GDP loss estimates may be overestimates, and there are some questionable assumptions. Therefore I wouldn’t put much stock in the actual results, but I do like it as a better attempt at methods and much better than merely looking at death rates given an assigned statistical value of a life.
And Now a Word From the Lamestream Media
The media is being impacted rather directly: “Roughly 36,000 workers at news companies in the U.S. have been laid off, been furloughed or had their pay reduced. Some publications that rely on ads have shut down.”
~
But assuming the media survives long enough for a retrospective, can we ask what went wrong with the media’s coronavirus coverage? And can we do better? Writing for Recode, Peter Kafka argues that the issue was about properly communicating uncertainty and risk, a question of which experts to trust, and how to properly communicate what they were saying. While the media was wrong, in many cases the experts were fairly wrong too. (This is why the only place you can truly trust is LessWrong.)
Scott Alexander writes “A Failure, but Not of Prediction”, arguing that predicting the spread of COVID was very difficult but we need to get better about making clear recommendations that are the best given the uncertainty:
Vox (owner of Recode) Co-Founder Matt Yglesias has a rejoinder that maybe we can’t have nice things:
Rob Wiblin counters that we actually could’ve predicted things just fine (and Wiblin basically did):
Don’t Forget About the Nonhumans!
Rethink Priorities researcher Daniela R. Waldhorn wrote a bunch about how nonhumans are being impacted:
~
Previously I wrote about a short, scientific e-book about pandemics and animal farming. That e-book is now available in a significantly prettier version.
Your Regular Dose of WTF
A Second Dose of WTF
Fun (Online) Distractions, Because We All Still Need to Enjoy Life
If “Friends” were in quarantine.
Here’s everything coming to Amazon Prime Video in May 2020.
Banksy’s Working From Home and Says His Wife Hates It
Looking for a little something to spice up your next meeting? Invite a llama or goat to your next video call. The pricing is definitely on the pricey side but could make it happen! “For $65, you get a 20-minute virtual tour of the farm for up to six call participants. For a bigger meeting, you can pay $100 for a 10-minute animal cameo or $250 for a 25-minute virtual tour.”
It’s now possible to get a virtual haircut: “How it works? Step 1: Get tools ready. Find or buy your best pair of haircutting scissors or razor for you men’s, women’s or kid’s haircut. Step 2: Book an appointment. Step 3: Video chat with your stylist who will coach you (or your friend) through your haircut session.”
A Seaside Irish Village Adopts Matt Damon: “Sightings of Mr. Damon have become common in recent weeks in Dalkey, a seaside resort town southeast of Dublin, where his presence has added yet another surreal layer to life under lockdown.”
Prickles the Sheep Returns Home After Seven Years: “After missing years of shears, the voluminous creature had ballooned to about five times the size of a typical sheep”
-
Today’s briefing was made possible with significant work by Peter Hurford, Derek Foster, Daniela Waldhorn, and Neil Dullaghan. This brief greatly draws upon reporting by Johns Hopkins, The Dispatch, FiveThirtyEight, STAT News, Foreign Policy Magazine, Politico, and others.