[majormajor]

So, summarizing the timeline here:

Bad things happened.

Review-and-approve regulation happened, things improved.

Another bad thing happened, at a much lower scale.

Regulators cracked down, professional admins replaced experts, things got buried in red tape.

The author argues against review-and-approve but they don't present a better alternative than that first review-and-approve. They mention, but don't really endorse, liability law models. But today's adversarial legal model in the US is also slow, expensive, and results in tons of cover-your-ass legalese at large places.

The "ratcheting up" of the regulation enforcement seems like the much bigger problem here than an "review and approve" model itself. The need to look like you're doing anything to try to prevent bad things from ever happening.

We don't see that everywhere - we enforce speeding and reckless driving laws, but sometimes people still die from traffic incidents caused by those. The reaction hasn't been to put ten times as many cop cars on the road, or ubiquitous speed cameras. We did put in cameras for red lights, but those lost a lot of momentum and were pulled back in some places instead of rapidly turning into constant monitoring of everything about driving.

So is the problem just that politicians don't relate or understand something like a medical study as well as they do driving, and don't understand the tradeoffs and burden required to try to make sure nothing went wrong, ever? If anything, I'd expect that to result in industry lobbyists pushing deregulation to have a much easier go of it than they have. So I don't get it.

[fallingknife]

I think a lot of it is timing, bias towards the status quo, and how much immediate tangible benefit there is to the consumer. Take the example of electricity. It's quite easy to build a coal power plant and extremely difficult to build a nuclear power plant even though the former is much dirtier and more dangerous than the latter. The reason is that when coal power plants were invented, the alternative was no electricity. It would have been politically impossible to make it difficult to build them or shut down existing plants. People would have rioted. Politicians would have been voted out. There were power plant accidents, horrific pollution, tens of thousands of deaths from coal mining disasters, but nobody is going to go back to not having power, so this was just accepted and became the status quo.

Then when nuclear power is invented and plants are built, the alternative is coal, not going without power. So when there are nuclear accidents, the regulators can go nuts and shut down plants, and prevent new ones from being built. Consumers don't care that much about what the source of their electricity is, and coal is the status quo, so no politician loses their seat for going back to it.

If nuclear power had been invented first, it and all its problems would have just been accepted as the cost of having electricity instead, and it would be extremely difficult to build coal plants.

[pjscott]

> They mention, but don't really endorse, liability law models. But today's adversarial legal model in the US is also slow, expensive, and results in tons of cover-your-ass legalese at large places.

They allude to no-fault workman's compensation laws, discussed at greater length in this excellent article:

https://rootsofprogress.org/history-of-factory-safety

The trick there was to take the liability out of the adversarial court system, by not requiring the injured party to prove that the employer was at fault -- only that they got injured somehow. And it worked much more reliably, with less ass-covering and less overhead, than the previous tort-based model.

[firstlink]

All three of these things are orthogonal:

1) Negligence vs strict liability

2) Tort model

3) adversarial truth-seeking

[pjscott]

Indeed, and I didn't mean to imply otherwise.

[youainti]

Thanks for the link, that was a fascinating read.

[Karrot_Kream]

> So is the problem just that politicians don't relate or understand something like a medical study as well as they do driving, and don't understand the tradeoffs and burden required to try to make sure nothing went wrong, ever? If anything, I'd expect that to result in industry lobbyists pushing deregulation to have a much easier go of it than they have. So I don't get it.

Yes, it's that politicians and importantly members of the public don't relate. Driving is so ubiquitous in the US that almost everyone does it. Almost everyone with the ability to decide things in America or who has the time to lobby for something has been a driver and can empathize with a driver who was doing their best but encountered something unexpected on the road and ended up in a crash (note my usage of passive language, specifically to build up this frame-of-mind, even though in reality it isn't like this.)

The average American doesn't see themselves as the creators of new drugs nor do they see themselves as airline pilots. Even if the harms American perpetuate via driving are actually worse than the harms of the other two, without the ability to empathize as the creator of a drug or a commercial airline pilot, they're much more likely to favor heavy censure for wrongdoing for something they see as being done by someone else. It's a case of "Look grandma was trying her best when she accidentally crashed into the kid" vs "Those elites at Johns Hopkins trying to pull the wool over our eyes." The less likely the average American is to work a job, the more likely the average American is at heavily regulating that job.

It's a big problem with American regulations because everyday harms are swept under the rug while rare harms perpetrated by a minority are dealt with harshly.

[harimau777]

Perhaps the difference with traffic regulations is that both the benefits and drawbacks of regulation are experienced directly by many regular people. Whereas the benefits and drawbacks of corporate regulation are most directly experienced by the upper levels of corporate leadership.

Granted, even then I'm not sure what the solution is. (Well, at least assuming that employee owned corporations are mostly a non-starter in the US.)