Of course, from a physics perspective, lower speeds means less force in an accident. To take an extreme example, if we set a nationwide speed limit at 1 mph (1.6 km/h in kangaroo units) and enforce it completely, we might not have any traffic fatalities. What if at a 5 mph speed limit, one person dies in total across the entire country? Do you say that driving at 1 mph is safe and driving at 5 mph is putting lives at risk? Well, I guess statistically that is the case, but only if you ignore all of the other factors that went into the crash and attribute everything to speed.

For most accidents there is more than one factor that contributed to the accident. While speed is a measurable and oft-cited factor in a crash, it is not necessarily the reason that the crash occurred. Example: Someone drives 75 in a 70 zone, stops paying attention, and drives off the road into a wall. Speed and inattention could likely both be cited by forensic examiners as a factor, but the crash may have had the same outcome regardless of the speeding and entirely determined by the inattention.

And to the other opposite, there are plenty of scenarios where someone could follow the legal speed limit of a road while operating their car at an unsafe speed. Someone could drive 65 in a 70 zone on an icy road, or past stopped traffic, etc. While they are following the legal speed limit, they are operating at a speed that puts others at risk.

Basically, speed limits are an easy to understand approximation of what is usually a good idea for most drivers in most cars under reasonable conditions. Given that, sometimes they are too high and sometimes they are too low.

You might find this page interesting:

https://www.ite.org/technical-resources/topics/speed-managem...

Higher speed makes all road behaviours riskier. I'm also from the Kangaroo states and while I know of not a single person who likes getting a speed fine I think the vast majority agree that having some speed limits makes things safer.

Australia doesn't have the same cultural views as America with regard to freedom. We have high taxes, virtually no guns, heavily enforced traffic policing, etc. On the Wikipedia page for road death per 100000 population Australia is about 5 compared to the US at 12. A large number of Americans seem to be culturally ok with this as long as the government doesn't interfere with their life. Hacker News has an international audience and you often see the different value systems at play in the comments.

I’m not arguing that speed doesn’t increase risk, nor that we shouldn’t have speed limits.

I’m arguing that they’re approximations of safe behavior. Prevailing traffic science suggests safe traveling speeds are heavily dependent on conditions, yet most speed limits are static.

I’m okay with the idea that people should not exceed the speed limit because we’ve chosen it as a reasonable compromise. I’m less okay with saying that those who break the speed limit are unsafe by default, or that those who follow the speed limit are driving at a safe speed.

In addition, hackers from the US probably tend to lean more libertarian than the general US population. By contrast, the hacker scene in e.g. Germany, while of course being government-critical, tends to lean more traditionally left

I consider my politics left-leaning and not libertarian.

I'm making an engineering argument about risk and blame, not a political argument about freedom. I am not arguing that we shouldn't have speed limits. Speed limits are like democracy. It's a terrible system... except for all of the others.

Ideally we'd have a better measure of unsafe movement -- We could much better measure risk if the speed limit was set to an equation based on the difference in momentum vectors between vehicles, the capabilities of the drivers and vehicles involved, the weather conditions, etc... but that has a much larger weakness in that we don't have a way to effectively measure/calculate/communicate those expectations to drivers.

I am all in favor of a utilitarian solution to the problem, but I'm under no illusion that it is a rigorous benchmark of risk.

The libertarian argument is that we shouldn't penalize risk at all. I am saying we should penalize risk, but that speed limits do a mediocre job of measuring it.

Fair enough, I was just adding to the parent's comment about different cultural viewpoints, but I failed to consider your own perspective on its own. I think your point of view is valid.

That said, to get back to one of your original examples: at least where I live, even if there is a general speed limit of, say, 80km/h, you could still get in trouble for driving that speed under adverse condition, like ice, heavy rain, heavy fog or darkness. I think the rule is that you have to guarantee to be able to stop within half of the visible distance So it's not like the law is completely static either.