Comment by runako | Hacker News Reader

Comment by runako | Hacker News Reader

Parent Full thread

Has anybody modeled what this does to the overall system?

My impression is the system doesn't have all that much slack to begin with. And then to reduce all the major airports at the same time? And with the (current) expectation that next week will be worse?

Edit: this feels ripe for a simulator type of game. Assume X% of ATC walk off the job every week because they have to pay their bills and can't work ATC for free any longer. Assume Y% of TSA do likewise. Assume FAA increases acceptable fatality risk by Z% weekly. Give little sliders for X, Y, Z. See what conditions are required to let us make it to December 1.

More edit: It would be cool to compare this to natural shutdowns. For example, how does a 10% reduction overall affect traffic as compared to a given Nor'easter or hurricane or bomb cyclone?

More edit: give FAA the power to e.g. shut down airports and rapidly move & re-certify ATC on other airports, like regional triage. Maybe shut down Hobby and Austin and put everyone at IAH. Move to sectors, so there's a single airport operating in Texas and surrounding states, ATL in the southeast, etc. Game out how far in advance FAA needs to make all those calls in order to minimize fatalities. Game out what is the date after which air travel becomes less safe than driving. This could be like Railroad Tycoon, except from the regulator's perspective.

I would expect the airlines to ad hoc create a reduced master schedule in the interim until capacity is restored. They do this for major holidays, but many months in advance. Here they will be doing 72 hours in advance. Flights won't get "cancelled" they'll be NOOP (Not Operating) which is different. As an Ops Chief, this is heaven (while losing money). Tons of spare planes available. Lots of time to work on backlogged maintenance on the planes. Major headache is parking: it's not easy to have too many idle aircraft for a sizable carrier. Stowing them overnight becomes a choke-point.

That's a good point, the airlines know how to handle reduced system capacity. I think in my hypothetical game I am more interested in how does FAA game out what capacity to tell the airlines.

For example, assume ATC is still not being paid around Thanksgiving week. How many ATC are still coming to work, for free, with no assurance of receiving back pay, on a holiday week, with a second rent/mortgage payment due in a week? Planning around that seems much harder even than planning around a storm!

Who’s doing your modelling at the FAA? Aren’t they all furloughed or working while not paid?

> As an Ops Chief, this is heaven (while losing money).

Assuming the airline survives.

> Has anybody modeled what this does to the overall system?

At this point I'd be more concerned about safety than secondary effects so I think they are making the right call. At the same time the economic impact will be massive.

I'm interested in such a model in part because I am curious where the safety line is. There's a point after which it becomes reckless to fly in the US.

I'm not. The only way to satisfy your curiosity is to sacrifice a few hundred people.

That’s the point of the model, so that it doesn’t have to play out in reality.

I wonder if the FAA even has enough data on fatality to do something like increase the fatality risk by 0.1% or something. Commercial air travel is so ridiculously safe and has been for decades; would they even have enough of a sample to estimate that with any accuracy? Particularly around something like having fewer ATC available, since I'm guessing even fewer of the rare accidents have that as a cause?

This is one of those annoying safety paradoxes, subtly linked to the prevention paradox. The better you get at getting rid of the incidents the lower the rate, the longer it will take to gather data to improve further. There is a similar curve involved when it comes to debugging. High frequency bugs are easy to squash. Low frequency ones can be really hard.

Yes, but if your model is wrong people will die. Unless you mean to use the model only from a curiosity perspective and not to base actual decisions on.

I don't think GP is an official at the FAA or has any influence over those who work at the FAA. It's kinda strange that you'd assume it's not for curiosity and/or personal use.

It is you that is making assumptions here. I assumed nothing.

Yeah, further down in my post I talk about making it a game, it being "cool" to make comparisons etc. I seriously hope the people making decisions have better tools and more insight than my late-night noodling, or we are well past the point where it is no longer safe to fly.

The longer it goes on, I expect another statistic which is once the backpay check clears, people quit. Because this is bullshit. The backpay will not cover the financing costs of going into debt to cover bills, food, rent, mortgage. The government should foot that bill, but I don't think they will.

Back pay isn’t guaranteed, and Congressional leadership has gone back and forth about whether back pay is owed. I would put money on any back pay being partial at best. (I understand that this is waste, fraud, and abuse that is being cut?)

But yes, this is bullshit. We also should not have active duty military using soup kitchens abroad. But on these matters, my opinion is obviously different from that of most voters. Hopefully, voters will change their minds.

Seems it is the law.

https://www.nteu.org/blog/2025/11/03/ShutdownBackPay

In the last few months, that must be accompanied by the question of who is going to enforce it.

Those charged with enforcement of the relevant laws have waffled on whether or both they plan to issue back pay. Their decisions after the shutdown are the only operative factor, regardless of the letter of the law.

Reply on news.ycombinator.com