Friday, February 28, 2020

COVID-19 # 18 (Required Parts are Highlighted)

No changes in the CDC warning levels since Wednesday.

State Department made one change: downgrading South Korea from 2 to a 3 on their 1 to 4 scale.

The first cases in the U.S. that are not related to travel to China were announced on Wednesday night. There’s a dorm at UC Davis that is now quarantined. This is what I meant the other day when I posted about community infection. We also found out that this patient has been in intensive care since the 19th: a week before the news was released.

The CDC issued this cute graphic about what men should shave in order to get a facemask to fit properly.

This is cool. It plots Chinese provinces in columns, and days in rows, with color coding for the number of cases.

Obviously Hubei is still pretty bad. Also, in a lot of those other provinces, the quarantines were not ordered by Beijing, but strongly encouraged by them, and enforced by local authorities. It tends to confirm the view of the WHO that the outbreak is mostly being brought under control in China.

The graphic also points out that quarantine is not a magic tool. If it was, Hubei, where quarantines are tighter would be in better shape. Quarantines are mostly about slowing the spread of the disease. In the chart, you could of no quarantine as making it more likely that red in one cell leads to red in the following cells. Thus the handful of red cells in the second, third, and fourth columns weren't allowed to perpetuate much.

Very interesting. The Chinese word for SARS is "Feidian". Check this out:

Searches for SARS were constant until around December 30th. Searches for feidian started going up around December 15th. Someone knew something was happening. Here's the source working paper.

China is postponing some national exams that were to be held on April 11-12. It’s hard to reconcile a cancellation that far out with an outbreak that is coming under control.

Xinqi Su reports that this new information was quickly censored in China (not really sure why), but it’s still available here. Caijing, a Beijing based magazine, reported that expert teams from Beijing that went to Wuhan in January were lied to by local officials about hospital staff becoming infected. This delayed the conclusion announced last week that the virus was capable of aerosol transmission (basically, particles that float for a long time, rather than just getting sprayed when someone coughs or sneezes).

P.S. On a related note, police in Hong Kong rounded up a bunch of anti-government demonstrators today, based on past offenses. Might this be because the unofficial/local quarantines there make it easier to find people?

Still working on this until class time.

Wednesday, February 26, 2020

May As Well Post a Map of China So You Know Where Things Are


COVID-19 # 17 (Required Parts are Highlighted)

The CDC has updated its warning levels. So has the State Department (updates are highlighted in purple):
Country
State Department
Center for Disease Control and Prevention
China
4
3
Hong Kong and Macau
2
1
Japan
2
2
South Korea
2
3
Italy
2
2
Iran
4
2

If you’d told me at the beginning of the semester I’d have to research this stuff, I wouldn’t have believed you. The new thing I’ve found out is that State Department levels go from 1 to 4, and 1 is no travel risk at all. The CDC’s levels go from 1 to 3, and 1 is used for places with some risk (the CDC is part of the Department of Health and Human Services, not the Department of State).Anyway, I color coded these so that the bad ones are in green, the worse ones in yellow, and the worst ones in red.
The International Olympic Committee is considering cancelling the Tokyo Olympics this summer. They are going to watch how the epidemic evolved until late May, and then make a decision.
There’s new research out on quarantine effectiveness.
  • Effective and interesting: the travel ban inside Wuhan slowed down the epidemic’s growth in other cities, but only by about 3 days.
  • Effective: closing entertainment venues
  • Effective: banning public gatherings
  • Effective: closing public transportation within cities
  • Ineffective: banning public non-air transportation between cities
No idea why they did not look at air travel.
I missed this from a few days back. Iran was not expected to be one of the first places an outbreak would spread from China:
Uh oh.
Add Brazil to the list. The infection has now gotten to all continents. Johns Hopkins has a dashboard on the outbreak, putting a map and numbers in an easy to find location. The problem with getting a case in Brazil is that Carnival season (similar to our Mardi Gras) ended yesterday. It is believed that an annual public pot luck dinner (basically a tailgate party) in Wuhan in mid January, attended by about 40K people, was a significant factor in the outbreak.
An interesting feature that you can pick up from the Johns Hopkins dashboard is the extent to which the healthcare system in Hubei was overwhelmed. It is not a fatality rate, but another thing that we look at is the ratio of deaths to recoveries (basically, patients who died in care divided by those who were in care and checked out of the hospital). In Hubei it’s 12%. The next hardest hit province in China is neighboring Henan, and there it’s 2%. In rich Guangdong, it’s under 1%. But, distance is a problem: Heilongjiang, which is about as far as you can get from Hubei, doesn’t have much of an epidemic but it’s 5% rate suggests they’re not prepared for what’s coming.
The WHO is supported by countries that donate in proportion to their GDP. China is its second largest donor. There has been suspicion that the WHO will go lightly on China to preserve their funding. On Monday, the WHO announced that they do not think there are a significant number of undiagnosed cases in Hubei. Outside experts are not so sure. Either:
  • The WHO is right, but fewer cases means the fatality rate is worse than estimated, but also that the risk of further spread is low, or
  • The WHO is wrong, and the fatality rate is lower, but the risk of further spread is higher.
Outbreaks in Iran and Italy point towards the latter. Also, WHO representatives at press conferences on Monday did not wear face masks, but on Tuesday they did. Interesting. But Julia Belluz noted this important point from those press conferences: [the] “Main driver is not widespread community transmission” but rather transmission within households. She also linked to this chart from the WHO's press releases:

This does tend to support the WHO's position that the virus peaked in Wuhan a few weeks back, but that their ability to diagnose the illness was lagging behind by 10-15 days. It's also telling that they were getting thousands of onsets of symptoms each day before they really started testing at all. The whole article is interesting, especially for its comparisons to the SARS and MERS outbreaks.
OK. And here’s some Twelve Monkeys sh*t. Most cases in South Korea are linked to the Shincheonji Church of Jesus. This is widely regarded as a cult: members are not allowed to tell other family members they have joined. This church’s eschatology is that all its members will attain eternal life through a rapture like event when the church reaches a certain size (about where it is now). Think about this:
A senior health official in Daegu — the city that lies at the center of South Korea's coronavirus outbreak — confessed to being a follower of a controversial doomsday church cult after testing positive for the virus.
The unnamed official leads the Infection Preventive Medicine Department in the city …
Excuse my implied language, but WTF. For its part, the Church’s public position is now that they know they are part of the problem, they are cooperating fully with authorities; in particular, services have been shut down.
On a related note, Iran’s conflicts with the rest of the Moslem world are mostly because they practice the Shia sect rather than the predominant Sunni sect. Shia is known for being more apocalyptic than Sunni, and Iran’s leaders over the last 40 years have repeatedly stressed eschatological points as motivating their political actions. Is it PC in America in 2020 to note that an outbreak might be hard to fight in a place that already suspects we’re in the end times? As evidence of this note that religious shrines have not been closed in Iran, nor have public religious gatherings been discouraged. And then there was this:
At a news conference that was televised live on state-run channels, the country’s deputy health minister, Iraj Harirchi, was quick to deny accusations that there was a cover-up of deaths due to COVID-19, the disease caused by the coronavirus. He characterized the outbreak as a “national problem,” and said that it was “not the time for political confrontations.”
But the deputy health minister himself was confirmed as someone infected with the virus on February 25—one day after he appeared beside government spokesperson Ali Rabiei at a press briefing, sweating heavily.
Iran currently has a higher fatality rate amongst reported cases than Wuhan ever had. And there’s new research arguing that for Iran to display the numbers it has publicly released, there must currently be 4-50K infections there (that’s a 95% confidence interval). Today, 6 new countries neighboring Iran announced a total of 61 cases.
On the other hand, this might make you rethink whether the U.S. policy of maximum sanctions pressure on Iran was a good idea. There’s probably a screenplay in that story …
Obviously, there’s always finger pointing when something like this happens. Dali Yang has an op-ed in The Washington Post arguing that, while China has made big improvements in disease reporting since the SARS outbreak in 2003, they failed monumentally this time around. Most of the blame is placed with local officials in Wuhan and Hubei.
I pointed out in a required blog post last week that you need to have a very strong filter when you read news that a U.S. president ever cut funding for anything. Here’s why:


The tweeter is a fellow at the Center for Global Development. The CGD is one of the most influential DC area think tanks. It’s not particularly political, but it’s fairly obviously in line with Democratic and/or progressive positions.







Tuesday, February 25, 2020

The Gambler’s Fallacy: This Is Going to Come Up In Lectures About Forecasting Soon, So It Is Required Even If There’s No Macro In the Article

Short and readable article about the “gambler’s fallacy”.

This will come up in a few weeks in discussions about how to interpret real GDP data if it contains a unit root (other similar ideas are “integrated”, and “random walk”).

Monday, February 24, 2020

COVID-19 # 16 (Required Parts are Highlighted)

A team from the WHO made an official visit to Wuhan over the weekend. It was the first time they’ve been allowed in. During a press conference earlier today they announced that the decline in cases there is real and not due to fudging the data. They assert that the epidemic was worst there between January 23rd and February 2nd.  Patients report that care in Wuhan for those with SARS-CoV-2 has improved. WHO also reported a death rate of 2-4% in Wuhan, but under 1% elsewhere; this is consistent with facilities in Wuhan being unable to keep up with treatment.
WHO emphasized that they are not praising how China has performed. But, they emphasized that mass quarantine was effective. I’m not sure how we’re supposed to reconcile those.
U.S. travel warnings are getting hard to keep track of. They are coming from two different sources, the State Department and the CDC. I put them in a table”
Country
State Department
Center for Disease Control and Prevention
China
4
3
Hong Kong

1
Japan

2
South Korea

2
Italy

1
Iran
4
1
Iran’s level 4 is about politics, arrests, and detentions.
The State Department’s level for China is specifically about COVID-19. Do not know why the two numbers for China are different, but they are. But get this:
I can believe the unfair treatment part. But, there are basically no anti-epidemic measures at all in the U.S., so the whole announcement seems hokey.
Italy has rapidly developed a big outbreak. There are videos from Rome of people in Hazmat suits; problem is the known outbreak is about 200 miles to the north. ANSA, the leading newswire service within Italy reports that a good chunk of the northern part of the country is on some sort of lockdown; 12+ towns are definitely in quarantine, with about 100K population.
Italy is part of the Shengen Area; where customs free travel between European countries is permitted. No country has asked yet, but there is a provision for this pact to be suspended if need be. Austria did unilaterally tighten its border with Italy.
Official stories out of China are that the outbreak is coming under control. Hard to reconcile that with this phone interview from Wuhan, where the quarantine has gotten tighter. Authorities in Wuhan announced that foreigners would be allowed to travel out of the city; this move was quickly reversed.
You may be seeing a new phrase: community-acquired infection. That means they don't have a clue how someone got it. This is what they're worried about in Japan and Iran.
For the first time, China has announced the aerosol transmission is confirmed. Basically, that's through normal breathing rather than coughing/sneezing. Not good.
Hubei has shifted back to requiring a positive test result to be counted in the official numbers. This will make them go down. This is the 6th time they've changed how the numbers are counted. Gee, that's helpful.
In an unconfirmed video from Twitter, a doctor from a hospital in Hunan confirms that he counted 50 new patients the other day and submitted those numbers to health officials. In turn, they included only 1 case in the official count.
There is growing suspicion that a 14 day quarantine may not be sufficient to declare someone virus-free; 27 is the new number being considered.
This graphic may be helpful:
The bars getting larger in several countries is not good.
China postponed the People's Party Congress in Beijing. It was scheduled years in advance for early March. It has not been postponed in 43+ years. It has yet to be rescheduled. This event is China’s annual meeting of party functionaries at all levels of government, drawing about 5K people officially, plus their retinues.
Car sales in China are off 92% in first half of February.
North Korea denies they have infections. BUT, they’ve quarantined all foreigners, and citizens who have traveled abroad. There are reports that a North Korean official was executed after breaking quarantine.

















Friday, February 21, 2020

Search On the Internet

There’s two ways that you can do search.

  • Keyword
  • The other way … this doesn’t really have a name … directed search maybe … following the links … I have heard of similar things being called “following bread crumbs” or “drilling down”.

If you know what you want, keyword search works very well. If you don’t know what you want, it can also work very well. And it’s fast. But maybe not very accurate, and certainly lacking in quality control.

But if you know a little bit about something, the other kind of search often works best. It also works best if you’re aware that keyword search will get you volume rather than quality, and you’re willing to poke around until you find just what you need.

**************************************************

This came up in class in the context of finding government spending numbers in GDP.

  • You need to know where to go. A keyword search for “USA GDP” will get you to the BEA. But you have to know that this is the place that collects and publishes that data.
  • Once you’re on the BEA site, you run into the problem that sites increasingly want you to use their tools to drill down for what you want. In the case of the BEA, you have to go through the Data menu at the top, then through one of the links on the left, which in turn opens up sub-links to the right. In class,
  • I went through Data, then clicked GDP on the left and then again on the right.
  • The next issue that you come up with is separating press releases and reports from the actual numbers. Sometimes you want a report; these are usually PDF, HTML, or DOCX files. Sometimes you want data, and this is usually in Excel/XLSX format or CSV. In class I clicked on “Tables Only” because it said it has stuff in Excel. When you choose that, a file downloads.
  • Except that now you’re in the position of having to drill down some more in that Excel workbook.
  • Now you have to think about whether you want nominal or real, dollars vs growth rates, and so on.

Or, you could back up. Instead of clicking Data in the third bullet, you could

  • Click tools
  • Click interactive data
  • Click GDP & Personal Income
  • Click “Begin using the data”
  • Then choose the Section you’re interested in
  • For example, I might choose “Domestic Product and Income”
  • When you follow this route, you get to numbered tables. Table 1.1.5 is the short form of what I showed in class.

**************************************************

Obviously, this isn’t complete, and maybe it never can be. I think maybe the important part is that everyone uses keyword search, but maybe we need to understand better that it works best at the two extremes, but not so well in the middle.

Thursday, February 20, 2020

Technological Innovations

Cute trivia here — we’ll spend some time in April discussing the importance of technological innovations.

Everything has to be invented by someone. Larry Tesler died today.He invented a great many things in computer programming.

In the early 1970’s he worked at PARC. This was a Xerox unit. located far away from corporate headquarters in Rochester, that was allowed free rein to investigate and develop novel computing techniques. For example, the mouse was invented there.

Anyway, Tesler invented copy-and-paste. You use this everyday, but once upon a time someone had to figure how to do that first.

COVID-19 # 15 (Required Parts are Highlighted)

New paper out on fatality rates. Their estimate is that amongst those infected (not tested, or confirmed, infected whether detected or not) the fatality/infection rate is 0.9%, with a 95% confidence interval of 0.4 to 2.6%. The estimated fatality rate for the global 1918-19 Spanish Flu pandemic is 2.04% (that is only available as a point estimate, data collection wasn’t good enough to estimate a confidence interval).
Pharmaceuticals for the U.S. market that are made in China are inspected by the FDA at the factories in China. The FDA has pulled out those inspectors. Expect some drug shortages here in the spring.
China has cut their publicly announced prime lending rate. This is an expansionary policy. But it is an exceptionally tiny move. This seems odd, given that China has made exceptionally huge interest moves in the past.
There are reports that the first person to come down with the infection (“Patient Zero”) was bedridden and had no connection to the Huanan wet market. That would tend to point away from conspiracy theories focused on virology labs and escaped bats.
People have locked-down in South Korea’s 4th largest city, Daegu, after an outbreak related to a church service.
Not good:

Remember those 2 hospitals that were built in 10 days? A Tibetan dissident website has posted a video from a contractor who worked on the construction. He says “I don’t know of any hospitals where doors open one way.” There are worse parts. Interpret with care.
Was in Wuhan, I guess:

BTW: China’s government appears to have been insulted by the headline of the article. But, in American newspapers, reporters write articles, while editors write the headlines. The editors are ll in New York. So why expel the reporters from China?
There’s some interesting good news. An internet service named BlueDot forecasts outbreaks based on sifting through internet news reports. BlueDot signaled a problem in Wuhan on December 31. That beat the World Health Organization by 9 days. Even better, reporters have tracked down that less AI and more human systems in Boston and New York started sending out warnings on December 30th. For perspective, the late Dr. Li Wenliang was reaching out — but only to his classmates from med school — that he was seeing odd symptoms on December 30th.
Roughly 200 million children in China have been restricted to online classes.
Iran is starting to report a growing number of cases, and in many cities. We have some expectation that countries with higher real GDP per capita will have better medical systems that are in turn better able to contain outbreaks. But, check your Handbook: Iran’s per capita, measured either way is comparable to China’s. Whether we like the quarantine idea or not, I give China’s more serious government a lot more credit for being able to do one successfully than Iran’s oddly prioritized government. Iran could be a big problem going forward. Also, Lebanon reported its first case, related to someone who traveled from Iran. And … umm … Lebanon has refugee camps.
Critics of how Japan handled the cruise ship with infections suggest that their government was very concerned with keeping the virus from jumping from the ship into Japan, but not very concerned about keeping it from jumping from passenger to passenger. In China, at least, they seem to jump into action when the virus starts to infect multiple people in a quarantined location (like an apartment building).
There’s a bit of anti-Trump bias out there. We talked a bit in class on Wednesday about how the plane bringing Americans back from the cruise ship in Tokyo were placed on a commercial flight with a barrier between the infected and regular passengers. That’s true. Here’s the tradeoff: the CDC wanted the infected people kept isolated (which is a good thing), but the Trump White House said no to that. This is being painted as a bad thing. But think twice about that; they also have to be concerned about encouraging panic. Worry about Iran’s medical system, not Trump’s decisions.
This is from a very big name in international finance:

Do note the use of the word “transitory”. You’re going to be seeing that a lot in class for the rest of the semester, because it’s part of the jargon of understanding how macroeconomic and financial variable grow through time.
Here’s a new buzzword to be aware of: “gray rhino”:

I like that one. I would clarify it to: highly probable eventually, with low probability at any specific time”. That’s more like the peaceful rhino in a nature documentary deciding to charge David Attenborough’s Range Rover.
My virtual friend Ironman, writing at Political Calculations, has developed an epidemic simulation tool and applied it to the caseload on the cruise ship in Tokyo. He found that everyone on the ship would have been infected within about 60 days. An isolated cruise ship may be a good model for the spread of the virus in a larger closed environment, like the whole planet. Uh oh.
This is kind of gross. A big reason for the spread of SARS in 2002-3 was that it took the fecal-oral route (you can look that up, or imagine what it means). Now actually, this is what common cold viruses do … bathrooms are … hmmm … cleaner in some ways than you might imagine, and dirtier in other ways. Obviously, it’s a good practice to wash your hands. Also a good practice to not use the same bathroom as someone who is sick. In the case of SARS, a particular problem was improperly installed plumbing fixtures, which allowed the virus to travel from a toilet in one apartment to a toilet (and then into the air) in another. Anyway, doctors are starting to report evidence that SARS-CoV2 can infect the same way.
Posting these items when I get a chance, will be updated until class on Friday.

Wednesday, February 19, 2020

COVID-19 #14 (Required Parts are Highlighted)

It’s been 5 days since the last class, due to the long weekend. A lot has happened.
Here’s the Reddit thread I mentioned in class. Ostensibly it’s about cooking, but it’s mostly about life in China under quarantine. The comments are just as interesting. Apologies in advance for the language sometimes used. The writer is in Shunde, in Guangdong, about 500 miles south of Wuhan. I have been unable to find a current list of quarantined regions in China; some are imposed at the national level, but most are being imposed by party officials at the very local level.

That’s an educated guess as to the size of the Chinese quarantine, which comes from an analysis by The New York Times. There are many reports that, outside of Hubei, local party officials have some latitude within China’s “grid management system” to make their local quarantine more stringent or more lenient.
This is a cool visualization:

Clearly, Chinese air traffic is way down.
China has fired the bureaucrats in charge of Wuhan (city) and Hubei (its province); essentially a mayor and a governor.
For perspective, 92 kids have died of the flu in the U.S. this winter. Precise numbers are hard to get for seniors (who often die “with the flu rather than “of” the flu), but it’s in the thousands. Do not forget that COVID-19 is mostly about what it might do, not what it has done.
The article I mentioned in class about "China sacrificing a province to save the world" is here.
As of last week, the U.S. has set up 15 quarantine centers around the country. I believe 12 are still empty. About 600 people are in quarantine, although the first group was released last week.
Also, a cruise ship with almost 4K passengers and crew has been quarantined in Tokyo harbor for a few weeks after passengers tested positive. The quarantine has been effective in keeping the infection from spreading into Tokyo, but not from spreading amongst those quarantined.
Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases at the National Institutes of Health, said the original idea to keep people safely quarantined on the ship wasn't unreasonable. But even with the quarantine process on the ship, virus transmission still occurred.
The Japanese health ministry said Monday that the number of cases confirmed aboard the Diamond Princess had reached 454.

"The quarantine process failed," Fauci said. "I'd like to sugarcoat it and try to be diplomatic about it, but it failed. People were getting infected on that ship. Something went awry in the process of the quarantining on that ship. I don't know what it was, but a lot of people got infected on that ship."
This Japanese doctor who went on to the ship explains that the quarantine within the ship has been inadequate. He has worked in areas with Ebola and SARS, and was much more scared inside the ship:

He was asked to leave the ship after pointing out problems. The message here is probably that it’s been so long since anyone tried to execute a quarantine that no one knows how to pull one off.
The U.S. began repatriating people from that ship to quarantine centers on the U.S. mainland over the weekend.
This isn’t too surprising:

It would put the number of infected at about 500,000. That seems consistent with reports that Chinese hospitals have been refusing admittance to anyone not displaying severe symptoms.
Forecasting is always tough though, and predicting epidemics is especially hard. This article entitled “Disease modelers gaze into their computers to see the future of Covid-19, and it isn’t good” discusses forecasts made by the CDC. Their range currently runs from 550,000 to 4,400,000 infections. It’s by Sharon Begley, an award winning author who I’ve been reading since the 80’s.
Lastly, this piece, entitled “How China’s Incompetence Endangered the World”, appeared over the weekend in Foreign Policy. That is a regular media rather than scholarly outlet, but is a serious one. The author, Laurie Garrett, has won a Pulitzer Prize and two Polk Prizes for reporting on international health issues. You can follow her on Twitter at @Laurie_Garrett.
The Chinese leadership seems to be getting more sensitive to outside criticism. Today they expelled 3 reporters for the The Wall Street Journal in response to a February 3 article entitled “China Is the Real Sick Man of Asia”. I’ve read the article a couple of times now — it won’t ever be required for class — but it’s hard for me to see that it’s saying anything that a lot of other sources on the internet are saying: the epidemic is a big deal, the response from the Chinese government hasn’t been great, there probably will be economic consequences to both, and it’s not at all clear that the Chinese economy is strong from top to bottom to begin with.
Many institutions treat Hong Kong as a separate entity from the rest of China. Today, the CDC issued a warning about traveling to Hong Kong. This is the first place other than China for which such a warning has been issued. Hong Kong currently has somewhat fewer cases than Singapore, so this is probably a statement not so much about the progress of the epidemic, but of the availability of treatment supplies given that Hong Kong’s largest trading partner is China.
The fatality rate seems to be holding in the 2-3% range, still less than SARS or MERS. On the other hand COVID-19 is more easily transmitted. And still, the flu causes a lot more deaths, although its fatality rate is in the 0.15% range. Flu also seems to be much harder to transmit than COVID-19; people get the flu because it is endemic in the population of humans, not because it is transmitted easily. So the long-term threat with COVID-19 is that it becomes endemic. That threat is why we have quarantines.
Statistical evidence has emerged that China’s infection numbers and death numbers are too closely correlated for typical data. This suggests the numbers are being massaged before being released.



Measuring China's COVID-19 Recession

This may be the best real time estimate yet on what COVID-19 has done to the Chinese economy. China’s power plants run mostly on coal. China’s coal consumption appears to be down between 20 and 45%.

This is measured in days since the Chinese New Year, which fell on January 25 this year. So, they’re usually down for about 10 days after that, and this slowdown has stretched on for almost a month now.

To get that to GDP we need to know China’s energy elasticity. A plausible value for any country is around one, estimates from 15 years ago suggest 1.5 is more suitable for China. Here’s the back of the envelope calculation:

  • Choose a round number for China’s GDP like $20,000B/yr
  • Coal consumption is down 20% to 45%
  • The elasticity suggests a hit of 30% to 70% for GDP
  • That’s $6,000B/yr to $14,000B/yr if it’s a discrete jump. It isn’t, so looking at that typical slope showing recovery by about day 25 in most years, that slope suggests effects so far that are perhaps half of that as China built up to a sustained shortfall.
  • This shortfall is new and gradual, let’s say it’s about 1/25th of a year so far (about 2 weeks). That converts to a GDP loss of between $120B and $280B so far, or –0.6% to –1.4% of annual GDP in total.
  • China’s economy in 2020 is roughly the size of the U.S. economy in 2008-9. During the worst parts of that recession, the U.S. economy was off $20B in 2008 III, $85B in 2008 IV, and $45B in 2009 I.

All of these numbers are sketchy, but the suggest that the effect of COVID-19 on China over a few weeks is already comparable to what a large recession did to the U.S. in a few quarters.

Tuesday, February 18, 2020

Geographic Correlation from a Trucker

TL showed me this image he found in a Reddit thread. It shows, as a chloropleth (heat map), the travels of long-haul trucker over a 16 month period.
Do note that it tends to emphasize the interstates a bit too much.

But, it gets at the idea that most of where America is, is not the political boundaries of America that you'd see on a globe.

FWIW:  I recommend the thread. People post data visualizations on various topics they've made.

Sunday, February 16, 2020

Trump’s Budget and It’s News Coverage

The Trump White House released its federal budget proposal for fiscal year 2021 last week. (The fiscal year runs from October 1 to September 30).
Keep your eye on the ball here. The executive branch administers budgets that have been passed. It has no power over budgets that have not been passed (other than they can make their priorities clear). Deciding on the money is all about Congress.
Also, do not be in denial about this. All elected officials like government spending. It is their job to spend other peoples’ money. What’s not to like?? Here is the convention I use: both parties like to spend more, but the Democrats might like it a bit better, while both parties like to cut taxes, but the Republicans might like that a bit more.
Anyway, for about 2 generations now (Trump, Bush II, Bush I, and Reagan), the legacy media and Democrats have been selling the idea that various government programs have been severely cut by Republicans. Not so. Republicans in the executive branch propose this sort of thing, but both parties in Congress can usually agree to spend more on just about everything, just about every time.
******************************************************
This is also a good time to clarify two bits of the language of federal budgets and macroeconomics that get everyone (including economists) goofed up.
First, in the U.S., some spending is discretionary and some is non-discretionary.
Discretionary spending is chosen and voted for every year. It can be cut, but it usually isn’t. Only a quarter of federal spending is discretionary.
Non-discretionary spending is set by a formula. The formulas are written in to laws, and can only be changed by writing a law that removes the old one and replaces it with a new one. Doable, but not easy or common. The other 75% of the federal budget is non-discretionary.
Second, the name for all federal spending is outlays, but, the part that is measured in GDP is smaller than that.
Outlays include all spending including transfer payments. It’s just under $5T. But transfers are about 2/3 of that, and most of those are non-discretionary.
Government spending on goods and services is the fraction that’s in GDP. This is because it’s spending. When we measure spending we measure it by who makes the decision about what to spend. Most transfers go as checks to households, who decide what to spend it on. So it gets counted as consumption.
In any variation on Keynesian theory, including the bastardized-Keynesianism practiced by elected officials of all parties basically everywhere … government spending on goods and services is the only one that can help the economy. Transfer payments are about being sensitive and empathic, and also about buying votes. But no one who is a macroeconomic professional claims that transfers do anything for the macroeconomy. Politicians love to tell us about the programs they’ve created, and how they’ll help the macroeconomy, but if those programs don’t buy goods and services they won’t. There’s just no support for that. Unfortunately, it’s a bait and switch, and politically, it works.
******************************************************
So what about Trump’s budget? The Washington Post has a good article about this, with excellent graphics, entitled “What Trump Proposed In His 2021 Budget”. It shows that the Trump White House wants to cut a majority of stuff. And it shows you that of the budget items, defense is the largest already, and the White House wants to increase that a little. What it does not tell you is that they are only reporting on the 1/4th of the budget that’s discretionary. If you worry about the overall size of the federal budget (outlays) … they pulled a sleight of hand and slipped it right by you unnoticed.
But a lot of the discussion is at the level of the op-ed piece by Margot Sanger-Katz in The New York Times entitled "What’s in President Trump’s Fiscal 2021 Budget?" She focuses on outlays. But then says things like this:
The president’s plan includes about $2 trillion in cuts to safety net programs and student loan initiatives. Those reductions encompass new work requirements for Medicaid, federal housing assistance and food stamp recipients, which are estimated to cut nearly $300 billion in spending from the programs. The budget would also cut spending on federal disability insurance benefits by $70 billion and on student loan programs by $170 billion.
Ummm … no. Just no.
Honestly, I wonder if she snuck a rhetorical trick into that first sentence. As written, it is false. Change the order of the words from “… about $2 trillion in cuts to …” to “… cuts to about $2 trillion in …”, and it’s correct. Ooooh. That’s sneaky, eh?
******************************************************
So, what’s the reality? This graphic is just for science-ish spending, but it applies more broadly. Deep cuts are proposed, and Congress then passes big non-cuts:

The bottom line is that going forward you should keep your eye on actual data rather than media discussions of proposals. Watch what they do, not what they say.

Friday, February 14, 2020

Those Gas Maps I Showed In Class

The chloropleth (heat map) of gasoline prices I showed is from Gas Buddy. Here’s a direct link to the map, but you can also access it through the Gas Tools menu.

The post about seasonal variation in gas prices is here.

COVID-19 #13 (Still Optional)

There was a big jump in the number of cases in China this week, all coming out of Hubei province.
**********************************************
This is a good time to learn about false negatives and false positives.
In doing hypothesis testing, the null hypothesis is usually that something is not present in the data.
In some contexts, this is called a negative. Then a rejection of the null is a positive. A false negative is then when the test delivers a negative that should have been a positive. A false positive is when the test delivers a positive that should have been a negative.
All medical tests are actually statistical hypothesis tests. There’s been suspicion for a couple of weeks that the current test for SARS-CoV-2, the virus behind the COVID-19 outbreak, is delivering too many false negatives, especially in the early stages of the illness. So, too many people are being sent home, then they get sick, and by the time they get back to the hospital it may be too late for them.
In macro, this comes up in trying to forecast business cycle turning points. We have lots of indicators for this, but all the indicators yield so many false negatives and false positives that they’re basically useless.
**********************************************
Anyway, for Hubei only, China is now reporting clinical diagnoses. This means a patient goes to the doctor, and if their symptoms match they are diagnosed. It’s not clear if the tests are still being used or not.
You may want to click through that tweet and follow the thread, this is just part 1 of 8. So there was a big jump in the numbers, which are now up to 60K infected. Deaths are not affected by this jump, but continue to climb towards 1,500.
**********************************************
The New York Times radio show The Daily ran a podcast today entitled “Fear, Fury and the Coronavirus”. It juxtaposes China’s efforts to fight the outbreak with its central government’s efforts to contain public anger about how they’ve handled it.
**********************************************
The Epoch Times is a newspaper based in Taiwan. It is decidedly against China’s current government. Having said that, they may have an easier time getting first person reports from inside China. They have been interviewing managers of crematoriums in Wuhan:
… He’s been seeing about four to five times the usual workload.

Of the 127 bodies received by the crematorium on Feb. 3, eight were diagnosed with the virus, while 48 were suspected of having the illness, based on their death certificates …
Let’s back out some math.
  • If 127 is four to five times normal, then normal is 25-32 per day.
  • This suggests that 95 to 102 extra bodies every day are from COVID-19.
  • But, if only 8 of the bodies were officially diagnosed with the disease, this suggests that death rates are 12-13 times higher than publicly announced.
Form your own opinion about how truthful any of that is, or what the margin of error on my simple calculations is.
Also:
The official at the first-mentioned funeral home said that about 60 percent of the bodies come from private homes, while 38 percent are transported from hospitals.
That tends to confirm the videos available on the internet of vans cruising neighborhoods.


Thursday, February 13, 2020

PMI as a Leading Indicator

TL emailed me and asked:

What are you thoughts on using PMI indexes to measure economic conditions vs GDP? (See below video) https://m.youtube.com/watch?v=ydBzV9KTF0Y

**********************************************

So, what’s PMI (in the context of macroeconomics, there are other abbreviations too).

PMI is short for Purchasing Managers Index. It is produced by the Institute for Supply Management (ISM). It’s based on a survey of purchasing managers in manufacturing about how they feel about their supply chain: is it indicative of a stronger economy, or not? It’s an index that’s scaled so that when it is above 50 the economy is expected to expand, and to contract when it’s below fifty.

In principal, this is probably a good group to survey. They should have a feel for how the economy is performing at the wholesale level, a month or two before that hits the rest of the economy at the retail level. So it ought to be a decent leading indicator.

The video is informative about all of this. But I’d say it’s careful to note that it isn’t meant to think critically about PMI.

PMI would be useful for forecasting if it was a leading indicator. Leading indicators are variables that match the patter of business fluctuations in the overall economy, but which lead it in time. For example, when the leading indicator peaks, the economy is expected to peak shortly after.

**********************************************

I replied with:

I think PMI is worth close to zero. But not zero.

In order to be useful for forecasting it would need to lead the business cycle. In order to be useful to use alongside GDP it would need to be coincident with the business cycle.

If I go to Google and enter "is pmi a leading indicator" I get this https://www.google.com/search?client=firefox-b-1-d&q=is+pmi+a+leading+indicator But a quick look confirms these are all stock market tip sites that are probably selling something. That's not encouraging.

The Conference Board is the think tank that keeps track of the Index of Leading Indicators. They have a list and the PMI is not on it. Another index created by the same group (ISM) is on their list. So that would work better. Even so, leading indicators have a poor record of forecasting turning points. Better than nothing, but ... not much better.

The problem with using anything like this is false positives and false negatives. All indicators give a ton of them. Good ones give less.

This is not to say that early information on how the economy is doing wouldn't be useful, since GDP is produced with a lag. But experience suggests there are no magic ones.

And, there's a huge problem with people looking at the numbers after the fact and assessing whether they're useful. This is sometimes called a post hoc ergo propter hoc fallacy. They do this towards the end of of the video you linked to. The real test is always can you state what the indicator says about the economy before it happens, and then confirm that it does?

I added this later:

One other thing, if you go to Google Scholar, and put in purchasing managers index, you will find stuff, but none of it is very recent, and none of it is from the bigger journals.

P.S. When you do macro, you get pretty used to thinking a new thing you've discovered is going to work, only to have your hopes dashed. The world's just a big, complex, weird thing to predict.

Get used to that. It doesn’t mean that it isn’t fun or interesting. Nor does it mean that it isn’t useful, or that you can’t make a killing by being better at it than others.

But, what’s left unsaid is that many novices impression of what should be predictable is based on success they’ve had with predicting a few things in the past. But, what one learns with experience is that there are some things that are inherently less predictable. Short-run and medium-run business fluctuations are one of those.

You Should Read This (But It Won’t Be On Tests)

I have been able to track down why I am having trouble using images on this blog for class.

I have used a blog as an educational tool in this class for over 10 years, but they’ve decided to “improve” the internet.

I believe I have a workaround for the rest of the semester. It’s a pain in the butt, and makes it take twice as long for me to do blog posts, but I guess this is life in 2020.

Here are the problems:

  • I use Blogger for the class blog. Blogger is owned by Google. I use a third party editing tool to write posts, because frankly, Blogger isn’t actually very good at … um .. blogging. Anyway, here’s their policy since last July: “… Google has completely removed the ability for third-party apps to upload images to Blogger …”
  • But that isn’t the only problem. Google has also blocked “mixed content”. This is a nerdy way of saying that if an image I like on a website isn’t coded properly, and I used it on my blog, Google will block it. They claim this is a security issue, but to me it seems like a content ownership issue (if you don’t want other to use your image, just code it improperly).

Let me be perfectly clear about this. The internet is global.

In the U.S. we have a doctrine called “Fair Use”. This is a form of permissible free speech that goes back to 1710 in common law, and which was formally codified in U.S. law in 1976.

In short, Fair Use says that you can use small bits of other peoples’ copyrighted work in your own work, with attribution but without their prior approval, for illustrative purposes, provided you’re not making money off their stuff. Basically, you can quote other people.

Other countries don’t like it when content creators do this on the internet.

No one admits this publicly, but other countries are threatening internet companies with fines if they make available in those countries webpages that would be covered by Fair Use in the U.S.

*********************************************************

I believe that using my workaround, you should be able to see the content I want you to see. However, if there are things that won’t load on your browser, let me know. This is a problem that mostly affects images.

COVID-19 #12 (Required Parts are Highlighted)


There is always the possibility that these are fake news:
  • There are videos coming out of Wuhan of body collection vehicles making pickups in residential areas.This is plausible if the hospitals are full, as has been reported extensively for 2 weeks now.
  • There are also videos of body bags being used to contain the body of more than one child. This also seems plausible if there are shortages due to either use or lack or production and transportation due to quarantines.
This one is definitely real. The U.S. is evacuating its consulate in Hong Kong. It is believed this is the first time ever in response to disease. Public announcements do not show a lot of cases in Hong Kong, so you have to wonder why an evacuation. On the other hand, there are other U.S. consulates in China (in Beijing, Chengdu, Guangzhou, Shanghai, and Shenyang) where there are presumably more cases that were not shut down. And there’s always the possibility that the employees in Hong Kong pushed for this.
To clarify, the epidemic is COVID-19. The virus itself has been named SARS-CoV-2.

**********************************************************
Oil prices have started to show the effects of COVID-19. Local gas prices will soon follow.
Here’s a little bit on oil first. Most people don’t realize that there is very little oil storage relative to use. Most stuff comes out of the ground, is shipped to a refinery, processed, and sold for use to its ultimate consumer in a matter of weeks. The reason for this is obvious: oil is dangerous to store. Gas is even harder to store, so if it doesn’t get used it mostly needs to be burned off somewhere.
This is part of the reason the retail price of gasoline fluctuates so much. Our demand for it is relatively inelastic, but its supply shifts a lot.
China is, and has been for several years, the largest consumer of oil.
Digression A good economics student should be able to defuse a good bit of environmentalist mumbo-jumbo by thinking that through. The U.S. economy is larger than China’s, but China uses more oil. This should tell you that we are more efficient. That shouldn’t be surprising: top to bottom we’re a richer country, and one way managers turn revenue into profits is by reducing costs, often through efficiency improvements. That’s also why we pollute less than China does.The implication is that economic growth leads to a cleaner environment. This should be fairly obvious if you compare a rich neighborhood to a poor one, or a rich country to a poor one. But that observation seems to be foreign to many people without enough exposure to basic economics.
Anyway, China’s economy is somewhat shut down, and has been for a few weeks. And this means that global demand for oil has shifted left. This means that prices should/will drop, and quantity too.


Quarantine was imposed in Wuhan and much of Hubei province on January 20th. You can see that prices for this particular type of crude oil are down about 15%. I do not know what the little uptick on the right hand side is from (perhaps businesses in China beginning to buy more in advance of the quarantine hopefully being lifted, or alternatively, since stockpiles are always low relative to demand, perhaps the oil tanks in China have emptied out over a few weeks and need to be filled).
This has not translated yet clearly into retail gasoline prices in the U.S.


We’ve been in a bumpy decline in gas prices for about 9 months. I haven’t explored why that has happened (I’m just thinking about COVID-19). Anyway, you can definitely see that the decline has gotten steeper and smoother over the last month or so. Data like this is always a mix of different impulses and the propagation of those, but my guess is that most of that steep decline is due to COVID-19 in China.
Quantity is a little bit harder to track, and will take more time. International quantities are what’s really important, and states who control their oil industries don’t like to make that information public. We could look at just the U.S., but there’s a problem there: advances in horizontal drilling (mostly) and fracking (less so, and very overrated) are making production per well go up. So they are mostly shutting down rigs, even when demand is strong. Anyway, there is weekly data on rigs in operation from Baker Hughes, and I might expect this to decline in the coming weeks:



I should return to this data and take a look at in a month or so.

Tuesday, February 11, 2020

COVID-19 #11 (Still Optional)

Monday, February 10, 2020

COVID-19 #10 (Required Parts are Highlighted)

Some historical review …
Both in class, and in the media, we keep seeing analogies drawn between SARS and the new coronavirus, and also sometimes with MERS. Both SARS and MERS were once the “new coronaviruses”. Both are associated with human proximity to animal reservoirs, followed by human-to-human transmission.
SARS hit Hong Kong and surrounding areas in 2002-3. It then disappeared (probably evolving into something we don’t notice much). At that time, SARS was the biggest epidemic scare in decades. Ultimately there were about 8K cases, and 800 deaths, in about 30 countries. A 10% death rate is high for a coronavirus.
MERS is more chronic, and has been popping up since 2012, mostly in the Middle East, with a big cluster in South Korea related to travel from Saudi Arabia. MERS now seems more serious: while rare, it hasn’t gone away, and the fatality rate is about 35%. There have been about 2,500 cases over the last 8 years, hitting almost 30 countries.
For both of these, there is a strong possibility that the fatality rate is much lower. Since they have similar symptoms to a severe case of the common cold, it’s probable that there are a lot of people infected who don’t ever get tested.
Keep in mind that these things are not as common as seasonal influenza outbreaks, but are more deadly. In the U.S. alone, there have been 22 million cases of the flu so far this winter, but the fatality rate is about 0.05%.
************************************************
The new coronavirus is now much more widespread than either SARS or MERS, with about 40K confirmed cases. And it passed SARS death toll this past Saturday.
There are two new issues with the data.
  • China has decided to change their counting rules. Now, if you test positive, but show no symptoms, you will not counted as a confirmed case. This is problematic, particularly in Hubei, where all hospital beds are full, and those without symptoms are being sent home.
  • For several days now, new cases in China have been level at just over 3,000 per day, with about 10K new ones every 3 days. This is not mathematically consistent with growth through compounding (which is what infections do). There is a possibility that this means preventive measures are reducing infections. But, a much more likely explanation is that China has hit the limit of how many tests they can process per day.
Both of these will make the announced infection numbers trend upwards at a slower rate.
Also note that when China started recommending the widespread closing of factories and stores outside the quarantine area, they indicated February 9th as the likely day to re-open. Not many (including Apple’s Foxconn affiliate) are.
************************************************
This is a good time to start reviewing the economic impact of SARS. SARS was geographically focused like the new virus. MERS is less geographically focused, so we’re not sure how much damage it’s done.
There have been lots of casual estimates for the cost of SARS, but the scholarly gold standard is probably Lee and McKibbin [2004]. You don’t have to read this, but you should skim it and take a look at some of the tables (like Table 2.2) and charts (like Figure 2.4). In particular, pay attention to the language: things like negative shocks, and temporary vs. permanent shocks, are mainstays in macroeconomics that you’ll start learning about shortly.
Their conclusion is that the cost of SARS, mostly born by Hong Kong, is about $40B in current dollars at the time the paper was written. Data from FRED shows that Hong Kong’s GDP in current U.S. dollars was about $160B/yr at the time. Now, total costs of an epidemic are a stock variable, and GDP is a flow variable, so be careful about comparing those (for perspective, at roughly the same time, the damage from Hurricane Katrina was estimated at $400B).
Their Table 2.2 estimates the permanent effects on Hong Kong’s GDP at –3.2%, and on China’s at –2.3%. That’s a lot. The long-run effects on the U.S. economy from the 2007-9 recession are in that range.
************************************************
While browsing over the weekend, I came upon a version of the initial (in depth) article in The Wall Street Journal, entitled “New Virus Discovered by Chinese Scientists Investigating Pneumonia Outbreak” that trouble was brewing in China. Reading it is sobering: this was published less than 5 weeks ago:
Chinese scientists investigating a mystery illness that has sickened dozens in central China …
Yep … dozens.
The disease afflicting patients in Wuhan hasn’t been transmitted from human to human, and health-care workers have remained uninfected …
It is believed that something like 5 million people traveled through Wuhan in the days after this on their way to visit relatives for New Years celebrations. No quarantines were in place until 12 days after this article came out.
Health experts say one risk is that the disease could become a bigger threat as tens of millions of Chinese travel around the country during the Lunar New Year holidays …
Health authorities in Singapore and Hong Kong, cities that have direct flights from Wuhan, have issued alerts …
Similarities to other diseases were still speculative:
Researchers have determined that a large proportion of new infectious diseases in humans are transmitted via animals. Such illnesses are referred to as zoonoses. Two newer human coronaviruses, MERS-CoV and SARS-CoV, have been known to cause severe illness and death, according to the U.S. CDC.
The Wuhan strain is similar to bat coronaviruses that were a precursor to SARS …
K.Y. Yuen, chair professor of infectious diseases at the University of Hong Kong’s Faculty of Medicine said this:
Given the marked advances in hospital isolation facilities, infection-control training and laboratory diagnostic capabilities in the past two decades, it is unlikely that this outbreak will lead to a major 2003-like epidemic …
Of course, China has been pretty clear about this for a couple of weeks now: they have a large region in which healthcare facilities are overwhelmed.
************************************************
Estimates I mentioned in class last week of a total cost of $150B are based on extrapolating the 2-3% permanent effect of SARS to China’s contemporary $15,000B/yr economy. It now appears it’s going to be quite a bit larger than that.