In the medical world, this is typically quantified by the Therapeutic Index[1], which is basically how much margin there is between an effective (therapeutic) dose and a lethal dose. (Fentanyl and heroin have notoriously thin margins, while cannabis is famously wide.)
It's a big consideration in what's OTC vs Rx, and what guardrails (inpatient only, regular blood draws, etc) are needed.
There was a good "magic quadrant"-style plot of this against addictive potential, but I can't seem to find it just yet. (This [2] is close.)
It doesn't consider externalities, so it's not great for the "do drugs and you'll end up unemployed, and if you're unemployed you'll end up homeless, and homeless people are a cost on the community" class of arguments.
It's not on shelves because it's a precursor chemical to actual meth manufacturing, not because it's a dangerous drug (otherwise it'd be Rx-only, like the other meth-like drugs).
Neither ephedrine, pseudoephedrine, or phenylpropanolamine are the reversed-chirality methampetamine. "Regular" methamphetamine is dextromethamphetamine; the other version is levomethamphetamine.
This may be true in some situations, but in others, it can result in automatic suspension of your license, and/or they can just get a warrant for a blood test.
I've also seen people convicted because, despite several hours' delay in obtaining a BAC, they were able to extrapolate back to the time of the accident. (There is apparently some dispute among forensic professionals on the validity of this, which implies the extrapolated value could be higher than actual)
Originally it was thought that alcohol metabolism had zero order kinetics (alcohol concentration goes down over time linearly, irrespective of dose/concentration) and now there is evidence to suggest it's first order kinetics at least in some regimes (rate of disappearance depends on instantaneous concentration). So the practical implication is they need to do a few more blood draws over time to extrapolate concentration a few hours ago:
- Lazy (old model): Take one blood sample, assume all people metabolize alcohol at the same rate, add delay time * generic zero order constant
- Not lazy (old model): Take 2 blood samples to determine that person't zero order constant + add delay time*K to original sample
- New model: Take 3 blood draws, do a quadratic curve fit, determine original concentration from that
To speak only culturally (and not legally), the rationale is roughly parallel to the Federal-State relationship set out by the 10th Amendment ("The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.")
I've spent most of my life in "commonwealth" states (PA, MA, VA), which (legend has it) are founded on a similar principle: the local governments can do what they want unless the state explicitly decides it's a state-level responsibility.
It's supposed to be a bottom-up structure, where middle tiers (Counties) are more a pooling of resources for the towns than a middle-manager for the State. There's often lots of "regional councils" or associations rather than "branch offices" of state agencies you'd see in a top-down structure. There may also be a County-level government with more or less power, depending on the state and specific topic.
They often have jurisdictional lines that reflect their pooling of grant-writing resources rather than simple geographic regions (like FEMAs) [1,2]. In the absence of statewide guidelines, these organizations often set the standards (often because the state can't or won't).
That is, the intermediate jurisdiction is often a voluntary collaboration - usually among the more progressive/better-funded townships/departments - and it's often the small-town departments that don't participate in these councils that have this "leave us alone" attitude and especially lax standards.
It's a big consideration in what's OTC vs Rx, and what guardrails (inpatient only, regular blood draws, etc) are needed.
There was a good "magic quadrant"-style plot of this against addictive potential, but I can't seem to find it just yet. (This [2] is close.)
It doesn't consider externalities, so it's not great for the "do drugs and you'll end up unemployed, and if you're unemployed you'll end up homeless, and homeless people are a cost on the community" class of arguments.