There’s a functional difference if you’re trying to compute “tough-on-innocence statistics”, because it means that the target for the only possible metric is much higher than zero. A healthy criminal justice system should routinely rule people innocent; if nearly everyone who’s prosecuted is found guilty, that doesn’t mean prosecutors are being really picky about filing charges, it means the system doesn’t treat innocence as an acceptable outcome.
You'd be better off assessing that via analysis of laws and policies. That would show what could be abused and to what degree, even if it's abused by the people in power now. That's really the whole point of the analysis - to create laws and policies with better protections, so we should measure what we're looking to change.
Yes, I agree. Rather than talking about "tough-on-innocence statistics", we should discuss what concrete procedural reforms would help people like Ms. Smith avoid being mistreated by the justice system.
One reason we're not doing that, I suspect, is that it's hard to figure out what such a reform would look like. What specific additional step could the police have taken to determine that she didn't operate this account, without access to her private knowledge that she's never been to Arkansas and doesn't have the type of phone that was used to create it?
