[Question] Why have exposure notification apps been (mostly) discontinued?

[VipulNaik]

I recently noticed that exposure notifications for COVID-19 in the app that I had been using, called CA Notify, had stopped working (I had attended an event a few days prior, and was feeling ever-so-slightly unwell, so I thought to check to see if there were any exposure notifications from the event; anyway it looks like my ever-so-slight not-feeling-well was unrelated to the event and it self-recovered quickly).

After digging further, I found that most of the COVID-19 contact tracing apps stopped sending exposure notifications around May 11, 2023, when the COVID-19 pandemic was officially declared over. See e.g. https://​​www.cnbc.com/​​2023/​​05/​​12/​​covid-most-americans-wont-receive-exposure-alerts-on-phones.html

I think this is a bit … sad? I feel like contact tracing /​ exposure notifications is one of the kinds of precautionary measures that provides benefits at ~0 personal cost, unlike things like social distancing, masking, and vaccinations. In fact I would have liked exposure notifications to be expanded beyond just COVID-19 to cover more respiratory diseases and also include the ability to report new ailments before they spread widely. Seems like keeping an exposure notification system going and having people regularly use it to report infections would be good practice for both known and unknown ailments.

I hope that companies and health institutions are secretly working on some more robust longer-term solution such as this, but have not seen any evidence of it so far.

What are the reasons this might have been shut down? I couldn’t find much discussion online, so I tried to come up with my own ideas, and here they are; I am not personally convinced of any of them:

1. Maybe the cost of running the backends was very high for the institutions managing the app (the California government in case of CA Notify).

2. Maybe people think “ignorance is bliss” when it comes to getting notified of COVID exposures, insofar as after having received such notifications, they may be morally obliged to cancel meetings with other people, incurring financial and social costs of doing so. They’d rather not be notified and incur the increased risk of transmitting COVID.

3. Maybe the exposure notification system was considered a security risk or liability risk of some sort, so shutting it down was considered prudent as a way to reduce the risk of some major scandal. (I think, even if this is the case, it’s probably better to tweak and fix any such issues now than try to rush out another iteration during the next pandemic!)

4. Maybe there’s evidence/​data that the exposure notifications system didn’t work at all? Or that not receiving an exposure notification gave people false confidence that they don’t have COVID?

NOTE: Originally posted to Facebook at https://​​www.facebook.com/​​vipulnaik.r/​​posts/​​pfbid02Q1jFK5LB7KhGWhj2rEteud8QfFrB2UxhyZYbWxQzSZWrpigJSasJEX7mVZKYmBZPl but I didn’t get responses on Facebook and I think LessWrong might be a better venue for discussing this.

[Dagon]

It’s not clear that it was EVER effective in any decision-making. In the early days, there were so many false-positives (WA State) that I pretty much know nobody who cared. And so many false negatives that you had to test ANYWAY if you had any symptoms or any personally-known exposure.

The problem is the app gives NO indication of duration or intensity of exposure, nor of the timing of the person you’re exposed to.

Almost pure theater—caring about the appearance of safety, not actual safety. Good Riddance.

[Richard_Kennaway]

The apps are over because Covid is over. Not just “officially” over, but actually over. It’s a minor illness now, worth about as much fuss as colds and flu.

[Gordon Seidoh Worley]

Not an answer, but at the time this happened some friends were distraught and I pointed out that my experience with the apps, having received notifications at various times, was totally ineffective. Specifically, every time I received a notification that I was exposed it was ~1 week after the fact at which point I had already developed symptoms (as was the case once) or not (as was the case 2 times). Knowing in ~1 week was fine for the early strains that had 14 day incubation periods, but once later strains with 2-3 day incubation periods took over it was just irrelevant.

The data was also spotty. The time I caught COVID while the systems were in place my wife never got a notification that she had been exposed despite being around me frequently during and after I caught it (was verified with a PCR from the doctor, so I know the system in theory should know I had it).

So in general I got the impression that the system probably wasn’t able to actually reduce infection rates, probably not even on the margin, since the time to notification was too long to be relevant in any meaningful way.

[Shaun]

I think there’s a trust value to discontinuing such programs when no longer necessary. I know here, there was a lot of privacy concerns around this framework, and the possibility it was a backdoor into long-term tracking.

The most effective response to such concerns was to be as open and transparent as possible, and the discontinuation of such frameworks could be seen along the same lines—earning trust by not allowing scope creep to turn these into permanent programs.

Ultimately it’s cost-benefit. We allowed the state to take steps we wouldn’t usually allow, because at the time the benefit was worth the cost. If that benefit has significantly changed, it’s worth re-evaluating whether it’s still a cost we’re still willing to pay.

[VipulNaik]

Hmm, aren’t exposure notifications an opt-in program? I was never forced to get them—I chose to download and install the app and keep it on. The same way I choose to allow Google Maps to keep a record of my physical location.