Privacy and Coronavirus Apps
Privacy and Coronavirus Apps
Life isn’t about finding perfect solutions but making difficult trade-offs. So it is with the contact-tracing apps that are proliferating faster than I can keep up with.
The goal, of course, is the same for all these apps: not to replace, but to augment and assist human contact tracers in pinging people who’ve been near an infected person. Human beings can only name contacts they know; software can point out strangers who’ve been within aerosol range.
The success of any tracing technology, however, depends on a lot more than its skill at sending "exposure notifications”. Once you get a ping, can you also get easy access to a coronavirus test? Does a social worker follow up to help you quarantine, by arranging for groceries, for example? Do public health officials keep checking in? Based on my experience, Germany and Singapore are both pretty orderly places that do these things well.
But to succeed at actually tracing contacts, any technology must first achieve what economists call network effects, according to which the usefulness of an app increases with the number of people using it. If you’re the only one on the corona app, its utility is zero. If everybody’s on, it’s invaluable. To be reasonably useful, at least three in four people must log on, experts reckon.
So far no voluntary app in the world comes anywhere near that. Even in Singapore, only about one in five residents has been using TraceTogether, which was launched as early as March. In Germany, only about 42% are even considering downloading the new app, according to a poll. What, then, are the factors that might be impeding adoption?
One is the relative inconvenience of using an app. Both TraceTogether and Germany’s Corona App, for instance, use Bluetooth wireless technology to let phones communicate with one another. But keeping Bluetooth enabled drains batteries faster.
The bigger issue is privacy and the perceived creepiness in being surveilled. Here apps across the world vary widely. Are users completely anonymous to the system or can they, directly or indirectly, be identified (as they can in China and South Korea, for instance)? How much control do people retain over their data? And what data, beyond infection status, is being collected in the first place? (In China and South Korea this included payment information.) Finally, will the data eventually be deleted, or could it be used for some other purpose?
This is where the data cultures of Germany and Singapore diverge dramatically. As you’d expect, Germany has opted for maximum privacy on every question down the list. Not only can users opt out at any time, they also remain completely anonymous. They may find out that they’ve recently been within two meters of an infected person for more than 15 minutes, but they won’t know where that was. Being anonymous, they also can’t be contacted, so it’s up to them to call in and report their health status or get instructions for next steps. In short, the Corona App probably won’t be all that useful.
Singapore skewed the balance differently. Its app safeguards users’ privacy from snoopers and contacts (that is, random people who’ve been near them), but shares information with the health authorities. That’s because doing so is useful. Officials can get in touch with infected persons and, with luck, identify the "super spreaders” who account for most transmission of SARS-CoV-2.
But as adoption lagged behind hopes, Singapore too realized it had to iterate and improve. It’s now preparing to ditch smartphones altogether in favor of new and bespoke devices, called TraceTogether Tokens, that will be distributed to all residents and are deliberately limited in functionality. These could be worn in a pocket or on a belt, perhaps. They wouldn’t drain the user’s smartphone battery, and they wouldn’t know anything about the person except infection status and identity. Unlike "smart” phones, they’re reassuringly "dumb” by design.
For now, Singapore is strongly encouraging people to wear these tokens, but it could eventually make them mandatory. It is thereby going in the opposite direction from Germany in every way. Is that illiberal?
It is, but that’s just another trade-off. Locking us down in our homes deprived us of our liberty. So did banning us from traveling to certain places, from sending our children to school and from mingling with friends and relatives. Countries that tried to preserve more of these freedoms, like Sweden, have paid for it dearly in lost lives. Good, widely used apps are part of the alternative. We divulge information in return for the ability to move about more freely.
Freedom, it turns out, doesn’t always equal maximum individual choice. Whenever we finally get a safe vaccine, we may need to require use of it by law in order to achieve herd immunity. Until then, we may need to coax people to participate in technologically enhanced tracking and tracing for the sake of epidemiological network effects. We should only support such measures on the precondition that they’re well designed and governed by democratically enacted laws. But a modicum of coercion may be necessary to prevent a second wave now, or the next pandemic later.