Complex Numbers

Risk, Revisited

This one has been hanging around in my queue for a while, and I don't remember the point I was going to make about it. If there ever was one.

It doesn't help that a friend of mine smuggled me a bottle of delicious mead all the way from Utah, and I just drank it.

Our Worst Idea About “Safety”  
A concept that took hold in the ’70s has haunted everything from seat belts to masks—and it’s going to keep putting us in danger.

The concept many medical experts can’t seem to loosen their grip on is known as “risk compensation.” It’s an idea that comes from the study of road safety and posits that people adjust their behavior in response to perceived risk: the safer you feel, the more risks you’ll take.

As far as transportation engineering is concerned, yes, that is a thing.

If you’re driving on a precarious cliff-side road without guardrails, you’d probably drive more cautiously. But some proponents of the idea make a stronger claim: that guardrails cause so much reckless driving that any potential safety benefits of guardrails are offset or even reversed. Under this reasoning, a road with guardrails would cause more accidents than a road without guardrails. Guardrails aren’t helpful; they’re counterproductive.

When I was in England several years ago, I was walking along a trail next to a stream when I noticed that something seemed off. After a while, I figured out what it was: even though the trail was, for the most part, significantly higher than the stream, there was no railing, no fence, no barrier of any kind. Here in the US, there would have been something like that -- or else they would have put the trail somewhere else, or, because this is America and we do what we want, moved the stream.

The idea behind the railing is to keep people from accidentally falling into the stream. Without the railing, though, I think we had a greater sense of the risks involved and were therefore more careful.

That's the kind of thing this article is talking about.

But whenever risk compensation has been subjected to empirical scrutiny, the results are usually ambiguous, or the hypothesis fails spectacularly.

Okay. I'm not going to argue with the science, if indeed it's saying what they claim it's saying. But I've thought for a while now that things are just too safe -- at least in the US. There needs to be some risk, some idea of the consequences of, say, falling off a two-meter retaining wall.

But in 1975, University of Chicago economist Sam Peltzman elevated what might have remained armchair speculation to a powerful argument against safety regulations. Writing in the Journal of Political Economy, Peltzman hypothesized that 1960s-era federally mandated vehicle regulations such as seat belts were actually making the roads less safe because they encouraged so much reckless and careless driving. In his thinking, any safety advantage of the new regulations was being offset. He analyzed traffic accident data before and after the regulations and found that not only did the regulations fail to decrease fatal accidents, but traffic-related fatalities increased after regulatory action.

I vaguely remember some discussion about that at the time, but I was nine or ten years old, so I generally ignored it while riding unsecured on a tractor, swimming in questionable water, or learning to drive a truck without seatbelts. Yes, I learned the basics of driving before I hit puberty -- one dubious advantage of living on a farm.

Note how one always "hits" puberty, by the way. One never eases into it.

Decades of traffic data now leave little doubt that, overall, safety regulations have indeed reduced traffic-related fatalities.

The fun thing about science is that sometimes science gets it wrong. And then it self-corrects, but people are so attached to the first conclusion that they ignore the correction. This is not a problem with science. This is a problem with humans.

Anyway, like I said, I don't remember the point I originally had in mind, but it came up today so I'm sharing it. I've been thinking about risk management in general for a while now, and there is, as I think the article points out (after all that mead, I can't be sure), a difference between personal and public risk management.

Tomorrow is the last day of the year, so it's very likely I'll do something other than riffing off of an article from somewhere, opting instead to do a review of the year or possibly a crystal-ball prediction for 2022. Maybe both. I haven't decided yet and probably won't until like 11:45 tonight.

And as the following day is New Year's, don't expect a blog entry at the usual time, because I may be drunk now, but I plan to be completely danchu at that time.