Thread by @aaronsnow

One more short thread about how #COVIDAlert played out, in part to explain more explicitly why I don't think anyone was really "wrong" here, given their respective circumstances. But it's really worth understanding what we can and should do about those circumstances. So: (1/n)

(This thread will make more sense if you've read the original blog post and/or the previous thread on this topic, both here:

https://twitter.com/aaronsnow/status/1539725685589266432

)

1. It absolutely made sense for the federal government to develop and deploy an EN service, nationally, even before it could gauge to what extent provinces and territories would or wouldn't effectively be willing and able to close the one-time-key loop in large numbers.

A huge benefit of rapidly deliverable technology is that you can place a bet quickly, at low cost, and then see whether "if you build it, they will come." It's especially beneficial when uncertainty abounds and time is of the essence, both very much true in early/mid 2020.

As others have noted, it's vital in a government setting to have data-gathering and decision apparatus and policy to govern whether/when to pull the plug. And the incentives influencing both the political layer and the civil service are complex and need to be managed; e.g.,

it has to be ok to press Stop without all involved being branded with scarlet Fs for Failure. But in urgent moments, sometimes everyone (through no fault of their own) lacks time and/or sufficient foresight to plan these bits as thoughtfully as they might in slower times.

(@sboots wrote much more about this last set of points recently, from an only-slightly-different angle: sboots.ca/2022/05/17/eve… )

2. Just as it made sense for the feds to place the bet, it also made sense from the PTs' point of view, in that moment, not to invest in solving the app's last-mile problem. There was a lot of painful fed-prov back and forth about this at the working and senior executive levels,

as of course folks wanted (and it made sense) to try to exhaust all possibilities before giving up on a service that could potentially save lives. But see my previous thread about why things didn't shake out in favor of that investment.

3. In another world -- and in a future state, if we're paying attention now and investing accordingly -- provinces (and states) _could_ make this kind of thing work. It will take the leveling-up of service management and tech skills governments clearly (still) need,

and it will take stronger fed/sub-fed working-level ("engineer to engineer" and "service owner to service owner") relationship bonds than currently exist either in Canada or in the U.S.

This is a price we pay for governing urgent national issues with highly federated authorities and structures. If we're not going to change that, we need at minimum the compensatory relationships and skills; and we need to understand that it will continue to slow us down. (/EOF)