Latest COVID-research (V) – NBER Working Papers

This thread lists COVID-related papers recently published in the Working Papers series of the National Bureau of Economic Research (United States).

From the 27 July edition:

NOTE: The NBER Working Papers series publishes early findings of ongoing research to encourage discussion and collect suggestions for revisions. Papers are neither peer reviewed nor endorsed by the NBER Board of directors.

Contact Tracing Apps: A Technology in Need of Testing and Trust

by Augustina Baker of (USA)

The past few months have put the world on a red alert.

Almost all countries have been hit with cases of the novel coronavirus. A sense of panic spread as the daily numbers grew exponentially for a short period.

Right now, governments and healthcare workers are more on top of things. With that, though, the curve is not yet flattened.

That has birthed different concepts of how to eradicate the virus, one of which is the contact tracing apps.

The Tech Behind Contact Tracing Apps

Contract tracing apps are used to trace contacts digitally. What contacts, though?

Manual contact tracing has long been established as an evidence-based approach to defeating the spread of infectious diseases. The COVID-19 is not the first global outbreak, which has given the healthcare sector the time and practical experience to perfect manual contact tracing.

Under this program, healthcare workers seek to identify those that might have come in contact with a confirmed infected case so that they can be isolated.

The diseases that require such measures are highly infectious ones. If the contacts are not moving, they won’t be infecting others. When they aren’t infecting others, the numbers keep going down.

While areas like California, New York, and Massachusetts have thrown a lot of manpower behind manual contact tracing, a majority of regional and national administrations may be waiting on digital contact tracing to get the job done.

That is where these apps come in.

Using the Bluetooth and GPS function on smartphones, healthcare officers can identify those that a positive contact has been around in the recent past. All smartphones with these apps on them will interact with one another in the background, storing data that allows them to be used to map one another later. Should one of the smartphones belong to an infected person, it is used to identify all other smartphones that might have been in the same areas as this infected person.

The course of action from here could be to send a message to such contacts to observe self-isolation for 14 days to be sure they do not have the virus already. An alternate course of action would be the evacuation of said persons by healthcare officials to designated isolation centers where they could be better monitored.

The Flaws of Contact Tracing Apps

However, with exposed backdoors in hardware and the increasingly commonplace nature of facial recognition systems, people have lost trust in government’s use of surveillance technologies over the years.

An ExpressVPN survey on contact tracing apps shows that more than 80% of the American adult population are positive that the government will tap into the data generated by these apps. Not even the promise from Apple and Google that the data will be locally stored on user’s phones is dissuading that thought.

The problem with this mindset is that the adoption rate of the app is affected. For digital contact tracing to work, about 80% of all smart phone users need to download and use the app. Take out the numbers that do not trust the government and big tech companies, and the adoption rate drops well below the required threshold. Even Singapore, which has had a head start on this project, has yet to attain a 30% adoption rate.

Say government intrusion and privacy were no issues. The numbers still do not account for people in rural areas, and those who do not have smartphones. After all, to get those apps means using an Android or iOS device. Even for those who have one, it would have to be a relatively new unit so that it is up to the task.

Final Word

This is not the time to throw all our weight behind an untested system, which the public is not warming up to. There needs to be more time for testing, integration with the public, and reinforcement of privacy-focused promises. Else, contact tracing apps will just become another lofty idea.

Report of restrictions in Wenzhou

In the last week of January, Hubei/Wuhan was locked down, and in Wenzhou (which also had a small but significant number of cases, especially relative to the rest of China) the government discouraged group meals. Since that was the exact Chinese New Year holiday week, most people had actually planned to have a lot of group meals with family and friends. It was all completely cancelled. Movie theaters and restaurants all closed down. People were asked to stay at home and not have group gatherings. Even small groups getting together for mahjong for example, was also not allowed and could be broken up and confiscated by police. People who don’t comply can be arrested. At that time the few people who walked around outside all pretty much wore face masks.

Shortly thereafter, the policy came out that said people are required to stay at home. The control of the local government is at the level of residential complex. Initially a system of paper and pencil residency passes was given to each household. That piece of paper allows one member of each apartment to leave the complex no more than once every two days for the purpose of buying food (note that once every two days is not per person, but per household). Each time someone goes out, a box on the sheet is checked off. I think the purpose was specifically for buying food only, although no specific time or location limit was implemented. Most people probably just wanted to get the groceries and go back home immediately, given what was happening in Wuhan. Each time going in and out of the apartment complex, temperature was taken. Face mask became a requirement for everyone who goes outside.

Shortly after that, the electronic version of health pass was implemented. It is a QR code which is in the cell phone App called Alipay. To get one, you need to input your real name and ID information, as well as the address you are currently staying at, and also filling out a survey about your health condition and whether you encountered any unhealthy people so far recently. You then get a color QR code, Green for OK, and Yellow and Red means not OK. My understanding is that since the cell phone location can always be tracked, the health authority can technically track everywhere you go in the future. If you go somewhere where later a virus-infected individual was said to have been (for example), the code might dynamically turn from green to yellow, or red. Only if the code is green can a person go anywhere, including outside of the apartment complex. Even after the QR code was implemented, temperature check was still being conducted each time going in and out of the apartment complex (up until recently). We can always check our own QR code status using our cell phone at anytime.

Although we have not gone anywhere really over the past weeks, we have some friends who used the QR code for their necessary travels. Some people say that the use and strictness of the QR code varies by province. Zhejiang (where we are, also home of Alipay) is one of the strictest. You need the QR code to leave the apartment, get into the train station, board the train and so on, everywhere you go.

There is one more notable thing I observed, which is that every evening during the commercial break of the evening news, there is a several minutes long public health message for people to watch. In the earlier days it reminded people about the laws related to coronavirus. In the more recent days it is a cartoon with music that reminds people how to prevent virus transmission in the workplace, such as sitting far apart when eating, spraying disinfectant regularly, including in the trash can, wearing face mask, and so on.

Basic covid-19 information in Italian: details

Da Galeotti e Surico: una versione in lingua italiana delle loro diapositive e lezioni. Queste sono informazioni importanti sia per il grande pubblico che per i decisori.



  • L’Enigma di Covid-19: Introduzione (Video)
  • L’Enigma di Covid-19. Prima parte: fatti ed interpretazioni (Video/Slides)
  • L’Enigma di Covid-19. Seconda parte: epidemiologia per non-epidemiologi  (Video/Slides)
  • L’Enigma di Covid-19. Terza parte: economia per non-economisti (Video/Slides)
  • L’Enigma di Covid-19. Quarta parte: politiche sanitarie ed economiche (Video/Slides)

Need for labor market flexibility

Michele Belot on

Quick re-matching of workers and jobs

As the pandemic shook the labour market, many workers became idle almost from one day to the next. Taxi drivers, restaurant serving staff, hotel staff, etc. are examples of service providers that are put on hold. In contrast, other services, such as delivery services, have faced a substantial increase in demand and have a hard time meeting the increase in demand in the short run.

It is rare to see a need for such drastic immediate adjustments in the labour market. At the same time, there exists a number of platforms providing on-demand services that are very flexible and can adjust well to fast and large variations in demand (such as Uber for example). Unfortunately, these platforms tend to be service-specific (e.g. taxi or food delivery) while now what we need is more flexibility in the nature of services offered on demand. For example, those working in the Horeca sectors should be able to help with the preparation of supermarket deliveries.

It should be relatively straightforward for existing platforms to create a specific sub-platform for matching temporarily idle workers to employers who have faced increases in demands.

In addition, one could think of a way of matching volunteers (such as young people) to elderly people for basic services such as food delivery. The goal here would be to minimize the needs for the elderly to leave their home and visit places that are open to the public.

Who is at risk?

From Michele Belot:

The key role of essential workers in stopping the spread of the virus

While we are all becoming accustomed with recommendations to washing our hands regularly and to practicing social distancing, there is a concern that the virus can still spread widely through a minority of people who are in contact with a large number of members of the public and/or who handle goods that are passed on to others. Even if most of us comply with social distancing measures, the virus can continue to spread easily through a minority of people.

O*net, a US-based platform classifying professional occupations has developed a measure of the degree of exposure to diseases and infections. Of course, most of these professions are in the medical sector, where protective practices are already common.

There are however several other professions outside the medical sector that are likely to remain considered essential throughout the pandemic but where protective practices are less common and rigorous. Examples include:

  • Retail salespersons, concierges, restaurant staff, and cashiers.
  • Couriers and messengers
  • Home health aides
  • Workers involved in the preparation of deliveries of goods
  • Bank and post office receptionists
  • Plumbers, handymen
  • Police

It is imperative that those in these professions follow the WHO and professional guidelines, including wearing a protective mask, washing hands regularly, and between every interaction with members of the public. Wearing gloves without disinfecting them between interactions may protect the worker but will contribute to spread the virus to members of the public.

Cheap and effective interventions

From Jan Kulveit: important, functional and cheap interventions.

  • Compulsory wearing of masks: Unfortunately the debate about masks was led completely astray by discussion about whether they are effective as protection of the wearer, and by the attempts of many governments to prevent hoarding by spreading the idea that masks are not effective as personal protection. However, the main benefit of everyone wearing masks is completely different: the mask limits the transmission from infected individuals. What’s crucial this works even in the pre-symptomatic period, and for asymptomatic individuals. So the wearer mainly provides a public good. Additional benefits are in encouraging behavioral changes. Masks remind everyone of the need for distancing. They are also great for signalling: people who are not taking the situation seriously don’t wear masks, or don’t wear them properly, and make it easy for everyone avoid close proximity with them. For how to get 10M people wear a mask in 2-3 days by promoting DIY efforts read here.
  • Scalable contact tracing solution based on use of telecommunications data: Here is how it works. Anyone infected is interviewed by an operator to track their past contacts. The operator is using our Memory Map, which includes telco data, banking data, phone location history and other sources to help the patient remember their whereabouts and contacts. All with user’s consent and within GDPR. New suspected cases are then invited by SMS to be tested. We are already extending this further to more data sources, mobile apps etc. The system was launched here in Czechia in one week including the necessary emergency legislation and the setup of the call centre. It is packaged and can be rolled out to other countries.

Both of these interventions are very cheap and effective. They have been effectively used in several Asian countries. Outside of Asia the only place implementing them I’m aware of is Czechia and Israel, and possibly Slovakia.

A user guide of Covid-19

Andrea Galeotti and Paolo Surico and I felt the need to learn some tools to process this information and to understand and rationalise current health and economic policies. We created a narrative to highlight different trade-offs which are part of: a) every health policy that attempts to flatten the contagion curve of Covid-19, and b) every economic policy that attempts to flatten the ‘recession curve’ caused by Covid-19. We have developed, to share with you, A user guide of Covid-19. It consists of a short introduction and four mini-lectures (each with slides and a video):

• A user guide of Covid-19: introduction. [Video

• A user guide of Covid-19: part I: gathering facts. [VideoSlides

Reporting the facts that are available to us all, in a way that is concise, precise and accessible to a large audience.

• A user guide of Covid-19: part II: epidemiology for dummies. [VideoSlides]

Explaining, in simple but precise terms, the epidemiology models that are guiding governmental health policies. Discussing trade-offs in health policies that implement different degrees of social distancing.

• A user guide of Covid-19: part III: economics for dummies. [VideoSlides]

Explaining, in simple but precise terms, the economic mechanisms behind the new economic crisis due to Covid-19.

• A user guide of Covid-19: part IV: policy options. [VideoSlides]

Discussing health and economic policy options in view of the current situation.

A file containing a comprehensive set of teaching slides is here: Lecture of Covid-19. This is a rapidly evolving situation, so we are updating the lecture slides (almost) daily.