Your link doesn't seem to establish a connection between DUI enforcement and reduction in deaths FYI (unless I somehow missed it...). Driving deaths from DUI ma

Your link doesn't seem to establish a connection between DUI enforcement and reduction in deaths FYI (unless I somehow missed it...). Driving deaths from DUI may have decreased while states were enforcing DUI laws, but the causality isn't all that clear. Meanwhile, distracted driving has exploded since the arrival of the smartphone and made it even harder to understand the dynamics behind road safety. That said, I once read that the most important enforcement factors, with regard to actually reducing DUIs, were 1) increasing the likelihood of being caught and 2) minimizing the time between being caught and experiencing consequences. Notably, severity of punishment wasn't a significant factor. I know that states will differ in enforcement approaches, but the US states I've lived in haven't seemed at all interested in adopting a data-driven perspective - what I've seen has been a very low likelihood of DUI detection coupled with prosecution that can take a year or more for a routine DUI (let's not forget needlessly harsh punishments that are more likely to destroy one's family or finances than decrease the likelihood of future DUI...). The contentious nature of this topic seems to make it hard to find rational discussions online. I've found plenty of emotional anecdotes, but few evidence-based suggestions mentioned (aside from the obvious but unrealistic suggestion to make cars unnecessary...). I'm no expert, but the "traditional American" perspective - 'Drunk drivers are horrible people who will go to hell when they die. Let's just charge all of them with attempted murder and lock them up for good. [commented on reddit via phone app from driver's seat of car in left lane on I-95]' (my characterization) - seems completely useless, in terms of actual harm-reduction. What does seem clear is that drunk driving is caused by poor judgment and/or inadequate risk-assessment (its correlation with # miles driven could be a third factor worth mentioning). If we observe the two largest demographic groups among drunk drivers, young men and chronic alcoholics, it's clear that they share these characteristics - young men because their frontal cortex hasn't matured enough to reliably assess risk in general and chronic alcoholics because their constant drinking necessarily impedes good decision-making. If we're not developing solutions that target those particular deficiencies (i.e. not a small chance at having a harsh, but essentially abstract punishment a year or more in the future), I don't think we'll see real progress on this front. In any case, I certainly don't think bringing enhanced surveillance into privately-owned vehicles is the solution.