PANDEMIC | 3. German, Italian Death Rates

This is the third of a series of posts on comparative coronavirus death rates. See also posts of March 19 (#1, on Germany's low death rate), March 20 (#2, on infection rates). Virus Deaths Tracker (FT).

	Cases	Deaths	Rate, %
World	307,278	13,049	4.25
USA	26,747	340	1.27
Germany	22,364	84	0.38
Italy	53,578	4,825	9.01

March 22, 2020–As of Sunday morning in the Eastern United States, the novel coronavirus (SARS-CoV2) had 307,278 confirmed infection cases in 171 countries in the world, according to the Johns Hopkins interactive COVID-19 map. Of these cases, 13,049 patients died, for a global death rate or case-fatality rate of 4.25 percent, as shown in the top line of the data in the table above.

The United States had a lower death rate than the world total, as one might expect because we are at the tail end of the westward spread of the disease, so our peak will come later. Surprisingly, Germany's death rate is less than one-third of that of the United States, and Italy's is about seven times that of the United States. Italy's death rate is nearly 24 times that of Germany's–why? I speculated on Friday in Post #2 that countries that test early flatten the curve.

What Explains Germany's Low Death Rate?

Why is Germany's death rate so low? I asked this question three days ago in Post #1, and I suggested that rapid early testing was an explanation. Here is a more comprehensive answer from a German retired corporate executive to whom I sent an email asking for his views:

I am not an expert on the virus, but I have some explanations for the low death rate: Germany has benefitted from:  1. A well-distributed nationwide lab system.  2. A "best practice" testing approach. 3. Most important, an immediate start to testing. 4. Germany has health care that is well established nationwide, health insurance for all citizens, cost coverage by insurance companies starting from the test to hospital treatment and hospitals with ICUs even in smaller communities.  5. Currently, Germany's legislative leadership, an alliance between Christian Democrats (CDU) and Social Democrats (SPD), was able to make quick decisions about expanding the number of ICUs and taking emergency measures. 6. Strong leadership from Angela Merkel and Jens Spahn (Federal Minister of Heath), both from the CDU. Plus Olaf Scholz (Federal Minister of Finance) from the SPD. They agree that money is less important than human lives, and they have the strong support of the European Commission and even their political opposition in the German Bundestag. However: 1. Germans expect a rapid increase of confirmed infections and higher death rates in the next two to four weeks. The biggest problem is to stop social interaction among citizens beyond core families, to break the chain of infections and "flatten the curve." 2. Rules about curfews and isolation have been left to decision-makers at the state level. Angela Merkel and her Government will decide on Sunday whether additional 

A CNN analysis puts its money on nurses per 1,000 population as an indicator of the quality of health care, and reports that Germany's 13.2 per thousand nursing staff is an indicator of high-quality medical care. https://www.cnn.com/2020/03/24/opinions/germany-low-death-rate-for-coronavirus-sepkowitz/index.html.

Another role model for responding to the COVID-19 disease is South Korea. That story is told here: https://thebulletin.org/2020/03/south-korea-learned-its-successful-covid-19-strategy-from-a-previous-coronavirus-outbreak-mers/. Rep. Caroline Maloney (NY-12) asked at a hearing how South Korea was able to test 15,000 people per day. http://english.hani.co.kr/arti/english_edition/e_national/933773.html.

What Explains Italy's High Death Rate?

Looking at the other end of the spectrum, what explains Italy's high death rate? That might provide some guidance for what NOT to do. I looked for answers and one came in via email from the Baltimore Fishbowl, a lively newsletter. Cara Ober interviewed Irene Woodbury, an American living in Italy. I have excerpted, and lightly edited to save space, portions of the long mid-March interview that bear on the question.

Italy was slow to respond to warnings. Italy is a few weeks ahead of the United States in spread of the virus. Italy's peak is expected in the first week of April. The Italian government closed schools and universities on March 4. But life in Italian cities and towns continued much as usual, with people eating, drinking and socializing in public. Not till cases kept surging did the government lock down the country. By the time they locked down Italy, we were at 300 deaths. The lockdown was incremental, because people resisted a complete quarantine. The 20-year-olds were thinking, “I’ll just be fine, it’s just a flu.” Statistically, they will probably will be, but they pass the virus to other people that it kills.  Now 20-year-olds are on ventilators in the hospital, and they can no longer assume they will be fine. Italians are very social and have had a hard time isolating. When the government issued a new lockdown in places like Milan, everyone ran to the train station and went home to towns in the south they’re from. The idea of two weeks isolated in their apartments was unthinkable. What made a difference was seeing people dying every day. Now people are taking it seriously. It’s weird, because you have this invisible enemy. Everyone is wearing masks and you feel you’re in a war zone. Socially, it’s so strange – no shaking hands, keeping a distance, trying to be nice to people without interacting with them. 
Supplies and equipment were lacking. People are wearing masks, but the stores all ran out of them, especially the good ones with the filters. At the beginning, they tested everyone, and healthcare workers often. When they started running out of tests, they rationed them. Now you get the test only if you have symptoms or have travelled to certain places. It has been taking up to three days to get results, because of the overload. But Italy's availability of tests has been better than in the United States. So 10 to 15 percent end up in the hospital in need of ventilators and the hospitals are running out of beds. It’s like a war zone. They’re opening up 30 new beds a day, using shipping containers to make beds and ICU units. The lack of ventilators is a national crisis. One company in Italy makes them, a small, family-run company, and they have accepted a government request to make five times the number a month; they have stopped making everything else and sell directly and only to Italy. They have brought in soldiers to help. If some people don’t get a ventilator, their lungs collapse and they die. With a normal flu, people might need a ventilator for seven days, but with this coronavirus it can require 20 days. Hospitals in the south are not as well equipped and prepared. Health professionals are scared. Nurses on TV are saying it really is a war for them, with thirty new ICU patients a day and hardly anyone leaving. Regular operations at the hospital have been shut down. If you have tonsillitis and need surgery, you have to  wait a month. "Sorry—we have to dedicate all of our resources to this virus." Italy has an enormous elderly population.  New cases include so many older people. One reason is that the elderly are used to grocery shopping every day. In public spaces,  the only thing you can do is go to the grocery store. Everyone is there. This is probably why the death rate is higher in Italy than other places.

Postscript, April 9— Here's an article that attempts to summarize the factors that affect reports of COVID cases n different countries.

PANDEMIC | 1. Germany's Low Virus Death Rate

The following is the first in a series of posts on this topic. See also posts of March 20 (#2, on infection rates), March 21 (#3, includes Italy). Virus Deaths Tracker.

March 19, 2020 – The data in the table below show a column for death rate ("case fatality rate") with a global ratio of 4.1 deaths per 100 cases of the COVID-19 disease as of this morning, using the Johns Hopkins database (jhu.edu). It's not hard to see that Germany stands out for its low death rate, 0.2 per cent. What's behind these numbers, and are there any lessons we can learn from them?

Why Is German Death Rate So Low?

Information on testing rates might be a clue. But data on testing rates, posted by Oxford University's Martin School's OWID, exclude Germany. The reason may be that Germany's automated testing system is so widespread within the country and distributed to such a wide number of labs, that they can't keep accurate track of all the tests that are being done.

Country (Rank by Cases)	Rank by Death Rate	Cases	Deaths	Death Rate	Recovered	Tests	Tests /Million Pop.	Comment
World	Total	222,642	9,115	4.1	84,506			See updates at JHU.edu
Germany (5)	1	13,093	31	0.2	113	-	-	Automated community testing.
Korea, S. (8)	2	8,565	91	1.1	1,540	286,716	5,566	High community testing.
Switzerland (9)	3	3,067	33	1.1	15	4,000	461	Testing number as of March 7
USA (6)	4	9,415	150	1.6	106	41,552	124	Inadequate, late testing; testing only those with symptoms.
France (7)	5	9,058	243	2.7	12	11,071	168	Number of tests as of March 10
Netherlands (11)	6	2,056	59	2.9	2	6,000	349	Testing number as of March 7
UK (10)	7	2,644	103	3.9	67	50,442	749	Updated at least daily. Data show number of people tested.
China (1)	8	81,154	3,249	4.0	70,535	320,000	2,824	Tests as of Feb. 24, Guangdong
Spain (4)	9	15,014	640	4.3	1,081	-		No data on testing
Iran (3)	10	18,407	1,284	7.0	5,710	-		Questions about transparency
Italy (2)	11	27,980	2,158	7.7	4,025	148,657	2,514	Unclear if people or tests. World’s 2nd highest elderly pop.
All Other		32,189	1,074	3.3	1,300
Comment						Some test numbers are for people, some for specific tests (x2). Community testing includes non-symptomatic cases.
Source	This blog	Johns Hopkins University JHU jhu.edu		This blog	JHU	Oxford University Martin School, OWID, ourworldindata.org/coronavirus

What is the story? 

Germany was the first to design a test differently from other countries; it has a rapid testing system. Germany has therefore probably diagnosed a much larger proportion of those who have been infected, and also a higher proportion of their population.

One factor is the number of independent labs throughout the country. Christian Drosten, the director of the Institute for Virology at Berlin's Charite hospital says that this network received both technical information to conduct tests and the approval to bill for them in January, when case numbers in Germany were still in the single digits.

Germany's distributed system helped doctors to determine quick whether suspected cases actually involved the new virus or a common cold, which can have similar symptoms. Drosten said: “Other countries lost a month or even more time because of this,”  Lothar H. Wieler, head of the Robert Koch Institute, Germany's equivalent the the U.S. CDC, said something similar: “We identified people early on with our testing who are infected, but not seriously ill.”

Why Doesn't the United States Adopt the German Test? 

The Germans developed a test for the new coronavirus one week after the genetic sequence for the novel coronavirus first became available in January 2020. That test became the basis for the World Health Organization’s (WHO) test used in countries around the world, including South Korea. However, the Centers for Disease Control and Prevention (CDC) declined to use it. The German test adopted and disseminated by the WHO scanned for three specific viral genes. 

Countries around the world adopted the test. South Korea, for example, quickly started testing around 10,000 people each day and now has tested 250,000 people.

The CDC instead created its own test using the three different genes from the German test. But when state labs tried the test, some found false positive results. The CDC redesigned and remanufactured test kits, but this took time.

Multiple Problems Being Addressed Now

The United States reported its first confirmed case of COVID-19 on January 21. Eight weeks later, there are aren’t enough tests for the virus available. Here are the contributing problems:

• Systemic failure. “The system is not really geared to what we need right now,” said Anthony Fauci, director of the National Institutes of Allergy and Infectious Diseases, at a briefing last week before the Committee on House Oversight and Reform, chaired by Rep. Carolyn B. Maloney (D-NY12). 
• Cuts in CDC, planning for pandemics. Contributing factors reportedly were cuts in the CDC and the termination of the unit planning for pandemics in the White House.
• The PCR is time-consuming. Until last week, the number of tests that could be run per day in the United States was limited to approximately 7,000, because it just takes so long.  The polymerase chain reaction (PCR) test has been around for decades. A doctor swabs a patient’s nose or throat and sends the sample to a lab, which looks for snippets of the virus’s genetic material. The test, however, is too slow. A bitter commentator has described the technology to me as "ancient." 
• Point-of-care testing is required in a pandemic. No one wants to wait for a lab while they decide where to put patients in an overcrowded hospital. "We’re just not quite there yet,” said Catherine Klapperich, director of the Laboratory for Diagnostics and Global Healthcare Technologies at Boston University.
• Shortage of supplies. The PCR swabs are imported and are in short supply.
• Too few labs. For weeks in the US, tests for the virus were being done by hand. A lab technician mixes a patient sample with tiny portions of chemicals in tiny tubes. Contamination can ruin the test. The samples must be repeatedly brought up to high heat and back down in a process called thermocycling, done on a specific machine. It takes a few hours to get results back. Labs that can do the testing have needed special approval to run PCR tests, and permission could take months. (Some of the problems listed here are being fixed.)
• State of emergency slowed FDA approval. Normally, state public health labs and commercial labs would have been able to create their own PCR tests.  But the public health state of emergency meant that any new test had to be approved by the Food and Drug Administration (FDA). That slow process also contributed to delays. On February 29, the agency relaxed the rule.
• FDA approval for machines took time. Machines are available that can run PCR tests automatically. However, it took companies weeks to develop tests for those automated systems and for the FDA to approve them.

Technology for the new tests is still experimental. The PCR test is hard to bring to a patient’s bedside, because thermocycling needed to run the test needs a lot of electricity. Another approach, isothermal amplification, is simpler and can work at a fixed temperature. Dozens of companies are now developing point-of-care tests for the novel coronavirus. The Verge cites three promising approaches: 

• The University of California, San Diego is evaluating a system that they say could return results in one hour. 
• Mammoth Biosciences is developing a bedside test that would work similarly to a pregnancy test. 
• Cepheid is leveraging a flu detection technology to build a rapid novel coronavirus test.

The question is whether, after a late start, any of these tests can be ready to help stem this pandemic. We should ask the Germans to help us get this done.

LEHMAN | Did It Have to Fail?

Paulson, Bernanke, Geithner.

James B. Stewart in The NY Times today reports on a new study of the dark days of Lehman Brothers in September 2008.

Could Lehman have been saved from bankruptcy? Should it have been? Would the world thereby have been saved from the Great Recession and its globally destabilizing consequences?

The study's author is Laurence M. Ball, Chairman of the Economics Department at Johns Hopkins University. He presented his 214-page paper, "The Fed and Lehman Brothers", which took him four years to write, to a conference of economists in Cambridge, Mass.

The study makes, as I read the story, two main points. Despite what Treasury and Fed officials (i.e., Henry M. "Hank" Paulson Jr., Treasury Secretary; Fed Chairman Ben S. Bernanke; and NY Fed President Timothy F. Geithner) have said,

• Lehman Brothers could have been saved. Bernanke told the Financial Crisis Inquiry Commission in 2010 that Lehman's collateral was weak and saving it would have required breaking the law. Ball argues that is not true, and that Lehman's financial condition was never properly analyzed. The whole point of the creation of the Federal Reserve in 1913 was to "lean against the wind" and when panic hits, its job is to save the system. The officials of the time underestimated the consequences of not saving the system and we live with these consequences today. 
• Paulson called the shots. Bernanke at the Fed followed the lead of Treasury Secretary Paulson, who took charge of the situation and was the prime mover in promoting the decision to let Lehman fail, because he didn't want to be known as "Mr. Bailout". Paulson says that the decision was that of the Fed to make.

Ball's paper was supported in its general conclusions by Prof. David Romer at Berkeley and another professor at M.I.T. Other academics interviewed by The NY Times withheld their judgment.

NYC Tech Industry - How Big Is "It"? Why Is It Important?

Roosevelt Island Campus, NYC, under construction.

A recent study estimates that the New York City workplace includes 291,000 jobs in the city's tech "ecosystem", with 150,000 of the jobs being tech positions in non-tech industries.

Within the tech industries, 58,000 jobs are tech jobs and 83,000 are non-tech jobs, a total of 141,000 jobs.

The study, The NYC Tech Ecosystem by HR&A Advisors, also concludes that 44 percent of the tech jobs do not require a bachelor's degree.

The study takes the correct position that to estimate the number of jobs in the "tech industry", we should include not just companies like Google and Facebook. It also believes that the count should go beyond occupational titles like entrepreneurs, coders, and engineers.

We took an inclusive stance toward the definition of software and IT jobs in the estimates we did for the NYC Comptroller's 1999 report on NYC's Software and IT industry, but it was not as broad as that of the HR&A report. I was the chief economist for the Comptroller at the time of the report, and Henry Etzkowitz was a consultant to the report.

Adding together the NYC and suburban job figures, the Greater NYC total was 47,000 in the software and IT NAICS codes, for the year 1997. This is one-third of the number of payroll jobs in tech firms in the HR&A estimates.

NYC is second only to Silicon Valley in number of
private tech companies acquired in 2012.

As a benchmark for assessing these estimates, Con Edison's 1999 staff for information resources was more than twice the number of software staff (293 vs. 140).

For software firms servicing Con Ed, the ratio of software staff to other information resources was slightly more than five to one (108 vs. 19).

My interview with Etzkowitz a year ago got into the question of technology as an economic development engine. It prompted some private comments to me from David Hochman that I have obtained his permission to share.

He wonders about the use of the term "tech" and "technology" in connection with the driving of economic development. Using the widest possible definition of tech jobs may not necessarily be helpful. See his blog post on this topic. The implications of his line of thinking include the following:

• The broadest definition of technology may be less useful for economic development purposes than a narrower one. Counting too many non-tech jobs in tech industries may make it hard to focus economic-development support. 
• Including too many non-tech jobs doesn't help the NY Tech Council, for example,  define itself as a "technology" council. Lacking a clearer boundary, it is tied down to software, web site and mobile communications functions.
• NYCEDC could fall into a trap of competing with Silicon Valley or Route 128 on the size of the tech work force rather than the quality of its tech initiatives. The challenge should be to get all the wealth creation, job creation and distribution we can out of the assets NYC brings to the global economy.
• NYC's universities could similarly get lost in congratulating themselves on the number of their graduates in tech industries rather than on the extent of their entrepreneurial skills and the next-generation readiness of their graduates. NYC's universities are still not obviously in the front rank in the tech field. The Roosevelt Island experiment is exciting but has as yet borne limited fruit.

The best news is that in Bloomberg's third term the startup innovation providers moved closer to the big, primary driving sectors. Hochman wrote to me:

You are right - this was resisted at first. What changed is that NYCEDC embraced the synergy/symbiosis, whatever you want to call it, and decided to make that the heart of the marketing/promotional campaign. Bloomberg asked NYCEDC to deliver marketing sizzle, and they did so brilliantly, joined by the Partnership for NYC and the NYC Investment Fund. I get at some of this here.

Hochman wasn't aware of the innovation group that met regularly at the NYC Comptroller's Office in the 1990s. However, he did know about the "New York Academic Consortium" of tech transfer officers that Etzkowitz mentioned in my interview with him. It was for many years bogged down in discussions of techniques of deal maximization, rather than on the more basic question of maximizing economic development. As Hochman says:

That's why I separately convened a group with some tech transfer officers but also many other university office directors with different titles and duties such as director of economic development or director of incubation initiatives. For a time we were the only group in the city - that I was aware of - discussing the cultural barriers to success. Only in the last few years have our institutional tech transfer offices become interested in spinoff formation at all, and it's still not the primary driver in offices that bring in so much money from drug royalties.

While I am a fan of the Cornell/Technion initiative on Roosevelt Island, Hochman has a warning about the way in which it was put together:

I am a skeptic... and I certainly disagree with your (tongue-in-cheek) conclusion that you should have tied your policy initiative to a real-estate program. I think that was a very bad outcome. Some of this I hinted at at my initial comments on the applied sciences initiative, but I haven't done the serious commentary I've been planning since Cornell was announced, though I certainly agree with you on the importance of Manhattanville and the role of engineering generally. 

These economic-development issues are not new. Stuart Leslie and Robert Kargon, professors in the History of Science program at Johns Hopkins University, wrote a paper, “Selling Silicon Valley: Frederick Terman’s Model for Regional Advantage”, Business History Review, Volume 70 (Winter 1996), 435-472. Long before Bloomberg launched his program, the paper explores the origins of the Silicon Valley model for regional economic development, and some problems with its replicability. We covered many of these issues  in Chapters IX, X and XI of the 1999 Comptroller's Report.

Leslie and Kargon show how the Silicon Valley model worked. Frederick Terman, Stanford's provost, envisioned a partnership of academia and industry, and trained students to put it into effect. He cultivated an aggressive entrepreneurial culture.

Beginning in the 1960s, business groups set out to move the model to other places, building their own versions of Silicon Valley, in some cases enlisting the help of Terman and his disciples. The authors discuss several examples, including the New Jersey Institute of Science and Technology, led by Bell Labs; the Graduate Research Center of the Southwest and the SMU Foundaton for Science and Engineering in Dallas, Texas; and the Korea Advanced Institute of Science and Technology - of which the Korean example is rated the most successful. The paper shows both the importance of the contributions of local universities and the limits to their ability to generate startups in the absence of a local tech industry that is available to nurture the environment within which startups can thrive.

More recently, in Invented Edens, Robert Kargon has moved on to write, with Arthur Molella, about models of the techno-city, i.e., a planned city developed in conjunction with large industrial or technological enterprises. The authors map the concept as applied in Norris, Tenn., home of the Tennessee Valley Authority and Disney’s Celebration, Fla.

In the early 1990s, the Disney Development Company set out to spend $2.5 billion to develop 4,900 acres in the Reedy Creek Improvement District. The plan -  prepared by Cooper, Robertson & Partners and Robert A. M. Stern - combines advanced energy and communications concepts with extensive parks and trails. Celebration is designed to be the ultimate techno-city and was named the "New Community of the Year" in 2001 by the Urban Land Institute.

The Terman model has served Silicon Valley well. We need to learn all we can from it. But it is not the only path to innovation in technology. New York City and Celebration and other communities need to explore other avenues.