COVID-19 isn’t a slow-motion train wreck – it feels faster than that (what day is it?). Maybe it’s a half-speed wreck: slow enough to watch, to intervene, but you still need to run. The problem is that the wreck is in every direction. Cities and states not only have to build capacity for testing and treatment (often with inadequate resources), but also manage the dire economic impact of ongoing social distancing; a challenge that will roll on until there’s a vaccine sometime in 2021. All of which is to say: This is how bad policy and dangerous precedent will be set, one jurisdiction at a time. It won’t be Congress overreaching, broad and plain. It will be in the states and cities (where there are few state or local papers anymore with reporters sitting at councils and legislatures) that privacy will be chipped away. Not for any ill intent, but because it’s a train.
Imagine being a governor right now. Your state health department explains that it has flattened the curve, but given the lack of testing capacity and limited clinical data, there’s every reason to believe that a wholesale removal of social distancing measures will bring the disease roaring back. Meanwhile, small businesses are failing, unemployment grows every day, and lines at foodbanks are miles long. Voters are scared; the virus, yes, but also for their ability to feed their kids and pay the mortgage. Further, you know that even if you open the state, many will still be scared to go out, and high-risk groups and marginalized communities will be faced with losing everything if they stay home and don’t work and losing everything if they go out and get sick. The options are horrible.
Then, in walks tech with an idea: use contact tracing to selectively identify and isolate micro-hotspots where you can push testing and resources. This will allow you to open the state, but through close tracking you’ll be able to quickly respond to outbreaks, preserve resources, and get people safely back to work. Governors and mayors, who are desperate for a solution, are contracting with companies around the country to do just this: use location data for tracing infections and disease surveillance. Two problems: contact tracing doesn’t work like that, and disease surveillance is more complicated than visualizing dots on a map.
Not that the general idea isn’t sound. We agree strongly that ongoing social distancing measures must be managed carefully – which is to say, ramped up and down as required – by both geography and demography. But without careful coordination with scientists and integrated privacy protections, governors will inject systems into their states that will both violate privacy and fail to protect citizens. The mania for contact tracing seems more to do with building cool tools than actual science, and most disease surveillance models seem to be science-free visualizations that do nothing to integrate epidemiological metrics or protect privacy. The solution – as we’ve argued before – is simple: any disease surveillance and contact tracing program must be joint efforts between academia, private industry, and government. Anything less is far inadequate for the challenge at hand and far too likely to establish bad precedents with which we will have to live for a long time.