The Economist UK - 28.03.2020

(Frankie) #1

20 BriefingThe pandemic and the state The EconomistMarch 28th 2020


2

1

them and alerts the authorities. Leaving
quarantine without your phone can incur a
fine; in South Korea fines for breaking
quarantine are hefty, and will soon be ac-
companied by the threat of prison.
Phones need not just send data back to
the government; they can also pass data on
to third parties. China’s Health Check app,
developed by provincial governments and
run through portals in the ubiquitous pay-
ment apps Alipay and WeChat, takes self-
reported data about places visited and
symptoms to generate an identifying qr
code that is displayed in green, orange or
red, corresponding to free movement, sev-
en-day and 14-day quarantines. It is not
clear how accurate the system is, but Alipay
says people in more than 200 cities are now
using their Health Check status to move
more freely.

Cellular biology
A group of academics, developers and pub-
lic-health officials from the World Health
Organisation (who) and elsewhere are
building a similar who MyHealth app.
When reliable tests for immunity—wheth-
er gained through infection or, one day,
vaccination—become available, such doc-
umentation apps may be used to commu-
nicate their results in some places, too.
When it comes to helping with model-
ling and situational awareness, there is a
wealth of data. Phone companies know
roughly where all their mobile customers
are from what cell their phones are using.
And because advertisers will pay to tailor
ads, internet companies such as Byte-
dance, Facebook, Google and Tencent gath-
er scads of data about what their billions of
users are doing where. Modellers can use
data from both kinds of company to fine-
tune predictions of the spread of disease.
Governments can use the same data to
check how their policies are performing at
a district or city level. In Germany Deutsche
Telekom has provided data to the Robert
Koch Institute, the government’s public-
health agency, in an aggregated form which

does not identify individuals. The British
government is in talks with cell-phone car-
riers about similar data access. It could
simply require it: the Investigatory Powers
Act of 2016 gives it the power to take what-
ever data it wishes from any company
within its jurisdiction in order to fight the
virus, and to do so in secret. In practice, ne-
gotiation and openness make more sense.
The belief that personal data are being
passed to the government in secret could
erode exactly the sort of trust on which an
“all in it together” fight, as called for by Bo-
ris Johnson, the prime minister, depends.
Google, which may well have more in-
formation about where people are than any
other company around, says that it is ex-
ploring ways in which it could help model-
lers and governments with aggregated
data. One example could be helping health
authorities determine the impact of social
distancing using the sort of data that allow
Google Maps to tell users about how con-
gested streets or museums are.
Computational social scientists, who
use data from digital systems to study hu-
man behaviour, are mulling over other
ways that this kind of data might inform
and improve epidemiological models. One
problem with current models, says Sune

Lehmann of the University of Copenhagen,
is that they assume that people mix and in-
teract in a uniform manner; that passing a
friend and a stranger in the street is exactly
the same sort of interaction. His research
group has written machine-learning soft-
ware which can sift through historical re-
cords from mobile-phone providers to di-
agnose and explore how relationships
modulate such interactions. Applied to
current data this understanding might
show that interactions between friends in
coffee shops are not that important for the
spread of disease, but that the delivery of
packages is—or vice versa. During an ex-
tended pandemic, such information could,
if reliable, be a great help to policymakers
trying to keep bits of the economy running.
The use of data becomes most fraught
when it moves beyond modelling and in-
forming policy to the direct tracking of in-
dividuals in order to see from whom they
got the disease. Such contact-tracing can be
an important public-health tool. It also has
a resemblance to modern counter-terro-
rism tactics. “The technology to track and
trace already exists and is being used by
governments all around the world,” says
Mike Bracken, a partner at Public Digital, a
consultancy, and former boss of the British
government’s digital services. To what ex-
tent those capabilities are now part of the
fight against covid-19, no one will say.
One reason governments keep secret
the procedures and powers by which they
seize and make use of data is a concern that
informed enemies would thus evade them.
When it comes to public health, this is un-
convincing. Complex as it is by the stan-
dards of rna-based viruses, sars-cov-2 is
not going to change its behaviour because
of what the spooks are doing. But their ad-
versaries are not the only people that
spooks like to keep in the dark. Citizens
concerned with civil liberties fit the bill,
too. This is why Mr Bracken expects gov-
ernments not to be forthcoming about any
use they are making of such capabilities in
the fight against covid-19: to be frank
would, he says, “expose the power that gov-
ernments have very quickly”.
Apparently unworried about doing so,
on March 16th Israel’s government autho-
rised Shin Bet, the internal security service,
and the police to use their technical know-
how to track and access the mobile phones
of those who have been infected. Israel’s
High Court initially limited the powers;
after parliamentary oversight was estab-
lished, though, they were good to go.
South Korea, too, is using digital sys-
tems to ease the load on its human contact
tracers. At the beginning of the outbreak
the Korea Centers for Disease Control and
Prevention ran their requests for location
histories through the police, who used
their channels to data controllers to re-
trieve the required information. But the

Staying level

Sources: Johns Hopkins CSSE;The Economist

Confirmed covid-19 cases, since tenth case
2020, log scale

*Data for China begin at 548th case

10

100

1,

10,

100,

13 20
January February March

27 03 10 17 24 02 09 16 25

South Korea

China*

HongKong

Singapore

Taiwan

China records
its first death

Hubei
lockdown

Singapore bans
entry from China

Daegu cluster
discovered

27,000 tests
carried out

More people recover than
test positive for first time

Largest daily increase
in new cases (48,
mostly imported)

New cases inside China
fall below those outside

Premier Li Keqiang
announces that the virus
is controlled domestically

Taiwan activates
Central Epidemic
Command Centre Teachers and students banned from overseas travel

1

At home, grown
Population under full or partial lockdown, bn
2020

Source: Press reports

2

2.

2.

1.

1.

0.

0
Jan Feb Mar

48 cities and
four provinces
in China Italy

Spain

France, Malaysia

Britain

191m Americans
urged to stay
at home

India

Argentina, Jordan
Free download pdf