The graph is not paywalled, and that’s really most of it. You can use IA for the rest: https://web.archive.org/web/20231211115919/https://www.nytimes.com/2023/

The graph is not paywalled, and that’s really most of it. You can use IA for the rest: https://web.archive.org/web/20231211115919/https://www.nytimes.com/2023/12/11/briefing/us-traffic-deaths.html https://web.archive.org/web/20231211115919/https://www.nytim... The general point, however, is simple: neither the US nor our peer countries significantly changed licensing requirements in the last 10-20 years. We didn’t suddenly stop doing public safety campaigns, either, but the United States is the only country where the fatality rate suddenly started going back up. The licensing hypothesis could explain why our numbers are higher in absolute terms but not that stark, sudden increase in legality – even if we made the license exams much easier in the mid-2010s, that wouldn’t affect the majority of drivers who already had their licenses by then.