[asdfasgasdgasdg]

The twitter thread makes this exact point. The camp fire would likely have happened with or without this maintenance error on PG&E's part. This is because a tree fell on a different power line and started the Camp B fire, and this was determined not to be the result of a preventable error. Just that fact alone suggests that the main component of the problem is not PG&E's neglect.

[jvanderbot]

Do you believe we should not penalize preventable deaths just because non-preventable deaths also happen?

[tikhonj]

That's not the question here. Rather, it's about which causes contributed to the problem—blaming everything on the most proximate cause is convenient but doesn't reflect the situation.

An extreme analogy: if a car is built in a way that a fender-bender causes a fire[1], how much of that is on the person who cause the fender-bender? Fender-benders are preventable by better driving, but the consequences were massively exaggerated because the car was unsafe.

The PG&E example isn't this clear but if the severity of the fires is substantially caused by poor forest management—an empirical question to which I do not know the answer—then it's clear that PG&E should not shoulder the entire blame. We're still left with the thorny question of how much blame they deserve (certainly not none), but the rhetoric I see around it is using the company as a scapegoat to let the CA and federal government avoid taking any responsibility at all.

[1]: I don't know if the Ford Pinto was actually this bad, but this is a hypothetical example anyway :).

[NotSammyHagar]

There are quite likely many other towers with the same vulnerability, the same potential breakage. Those will cause other fires when they break. There are multiple reasons why they should address those very old towers that haven't had maintenance, because they'll cause other fires.

[jvanderbot]

There are laws about negligence and inspection for utilities. There should be laws about forest management, I suppose, but the law is what determines culpability, not the cofactor analysis of the actual problem root.

In the fender-bender example, I can see how you would structure your logic the way you have. As an example of how I and others might structure it: The car causes the accident when a wheel falls off, and the accident happens to happen where there was a prior gas spill. The presence of the gas spill is unfortunate, but the car caused the accident, and that should never happen with good design and maintenance. The accident was exacerbated by the presence of gas, causing a fire and injury and death, etc.

Now, to bring it full circle. Replace a car with a train (the situation is largely under the control of a single company), the accident with a train falling off the rails and hitting a gas pipeline, AND the company knew the gas pipeline was there, and knew the danger all along, and still managed to not put on a proper maintenance plan to ensure the train didn't derail especially in a place that has a gas pipeline right there.

[asdfasgasdgasdg]

> but the law is what determines culpability, not the cofactor analysis of the actual problem root.

We are talking about what ought to be, not what is. The law falls under the category of is, and does not imply ought. I think everyone in the conversation understands that PG&E bears some amount of legal liability for these fires.

[asdfasgasdgasdg]

I think we should talk about how to design systems such that preventable deaths happen less often (supposing the costs to prevent them are appropriately scaled). I think hitting PG&E with arbitrarily large fines in this situation will not ultimately have much beneficial effect on the rate of preventable deaths. I think that their negligence happened to be the spark, but if not something else would have caused it. Like a tree falling on a line, a lightning strike, or a gender reveal party. Even if the fine led to perfect compliance on their part (it wouldn't/won't), that would not substantially reduce the rate of preventable deaths. This is because their negligence is not the central cause that most significantly raises the probability of large fires.