> The next thing you learn is that we really need to do something about tasers, because if there's a top 3 set of circumstances for death involving police offic

> The next thing you learn is that we really need to do something about tasers, because if there's a top 3 set of circumstances for death involving police officers, "death after tasing" is at least #3. Since you've looked at the data, could this be survivorship bias? Is there data on how often tasers are used where firearms would've been, in the absence of tasers? It certainly could be that tasers reduce the number of people killed by police, but it sounds hard to measure that.