Is 60% Even Possible?

What if they made a privacy-protecting Covid-19 tracking app and nobody used it? Or at least not enough people used it to make a difference?

NHSX is now testing a Covid-19 Contact Tracing App on the Isle of Wight, with a view to rolling it out across the UK. This app does not appear to satisfy privacy and digital rights groups, notably because it does not feature a de-centralised system for storing data. This is gearing up to be one of the major civil liberties battles of the early Covid-19 pandemic. There are a lot of interesting dynamics here. For example, the power that Apple and Google have, through control of smartphone operating systems, to dictate what software governments can, and cannot, build. The primary division worldwide is between systems that are decentralised by design (protecting user privacy), and those that are not, with the NHSX app firmly in the latter camp.

What has gotten me thinking, however, is the amount of users required in order for these apps to make a difference, and whether it is feasible for a liberal society (such as the UK) to achieve the number of downloads/stable users necessary for these apps to make a difference.

The headline figure, from Dr Professor Christophe Fraser, is 60% of the UK population (apparently this comes from the logic of the app working like herd immunity - see this early policy paper - also in detail in this Science paper). This could either refer to 60% of the total population, or, charitably, 60% of adults. Either way, that's a lot of people. If the target is 60% of everyone, then according to the Office of National Statistics’ projections for 2018 that's 60% of 66.4 million people (more or less), or 39.8 million people. If it's 60% of all adults (18+), that's 60% of 52.3 million people, or 31.38 million people.

But there's three problems: 1) There's a substantial fraction of adults who don't use smartphones, 2) Smartphone use declines by age, 3) Older people are a substantial fraction of the adult population in the UK.

According to OFCOM's Technology Tracker, 98% of people over 16+ uses a mobile phone, and 84% of those use a smartphone. However this declines dramatically at 55+ - only 64% of those aged over 55 use a smartphone. Unfortunately, according to ONS stats, this is about 20 million adults. Comparing OFCOM rates of smartphone use and ONS stats, there are about 41.4 million 18+ smartphone users in the UK. This is a back of the envelope analysis based upon published data, so I could be quite wrong here, but not wrong enough for it to matter for the next point: Even if I am out by a million potential smartphone users, we're still talking in the ballpark of 96% of all smartphone users downloading the app (if 60% of total UK population has to use it), or about 76% of all smartphone users downloading the app (if we need 60% of the 18+ UK population to use it).

Those are really, really, big numbers. This is YouTube, Facebook, Facebook Messenger, WhatsApp scale download and usage. Even if we were only talking about 60% of smartphone users, it would still be in the kind of app-usage numbers only seen by a handful of global tech companies.

So, how do you get there? In a liberal society, I'm not sure we do unless there are carrots and sticks attached to the app, and that's a big issue in liberal democracies for a whole host of reasons.