4 ms·

If you are arguing against overly stringent safety regulation by pointing out that they actually decrease safety, you have already lose the war against safetyis

If you are arguing against overly stringent safety regulation by pointing out that they actually decrease safety, you have already lose the war against safetyism. We must reach the point where it is widely permissible to say "the safety benefits are not worth the costs".

ChuckMcM 3y ago
And yet the FAA makes that argument explicitly in the article.
- jessriedel 3y ago
  That's my point. The FAA makes the argument, and Eli Dourado should reject the framing (but doesn't). There are no flying cars in a future where the main arguments for reduced regulation is increasing safety. It doesn't take you remotely far enough.
  - ChuckMcM 3y ago
    Fair enough, but that argument was not "for all types of aircraft" right? Are you arguing for the FAA to remove all safety requirements that have a cost? How do you evaluate that cost? When I read it I saw the FAA describing a methodology and argument that counters the "safety at any cost" argument (is that what you call 'safetyism'?) I read the article to say, "Sometimes our rules do not increase safety, sometimes our rules result in people choosing a less safe option because there isn't safer choice, and sometimes a new thing we haven't seen can be safer than what existing practice would expect." And MOSAIC was a response to addressing those observations in order to foster a more innovative and safer overall aviation industry. When I read your original point I mistakenly thought you were criticizing the FAA for lowering the cost of safety regulation when doing so increased safety. So that is how I got confused by your comment.
    - jessriedel 3y ago
      (No worries about the miscommunication.) > Are you arguing for the FAA to remove all safety requirements that have a cost? No. > How do you evaluate that cost? Lots to say here, but let me lead with a general principle: If you can measure the benefits of a regulation but cannot measure the costs, you should err on the side of less regulation. The original sin of cost-benefit analysis is "I can't easily measure contribution Y, so let's assume it rounds to zero". Here's one possible mechanism: Instead of banning an activity, either (a) put a tax on it equal to (say) 50% of dollar value of the safety harm (using, e.g., statistical value of life) or (b) require the consumer to pay for insurance that pays out a similar amount in case of accident. Then, even if we think the consumer is completely ignorant of the safety costs, they get to weigh the benefits to them of the activity against the financial costs from the tax/insurance, as determined by the regulators. Remember, the general intellectual justification of safety regulation is "We, the expert regulators, know the safety risks, but the consumer doesn't have enough information. Also, by typical paternalism arguments, it would be insufficient to simply create a Would-have-banned label ( https://mason.gmu.edu/~rhanson/wouldhavebanned.html https://mason.gmu.edu/~rhanson/wouldhavebanned.html ). It's very hard for us to estimate the cost of the regulation to the consumer, but based on eyeballing it, it seems smaller than the safety benefits." By taxing the activity at the level the regulators think, in their expertise, reflects the safety cost of the activity, they can let the consumer (who knows the benefits of the activity better) to decide whether the costs outweigh the benefits. Now, maybe if you implement that mechanism in the real world it turns out to be good or it turns out to be bad. Complicated empirical question. But the over-arching point is that no one even brings things like that up as possibilities. Which is evidence that the alleged justifications for the rules are not the real motivations. (Needless to say, the above suggestion is not new or invented by me; the ideas are just lying out there.) > When I read it I saw the FAA describing a methodology and argument that counters the "safety at any cost" argument It doesn't! It accepts the "safety at any cost" argument and then argues that, counter-intuitively, increased regulation reduces safety in this particular case. It's of course true that if a safety regulation both has costs and reduces safety then it should be removed -- it has no redeeming features. But we should be cognizant that this is an unusually easy situation. > the "safety at any cost" argument (is that what you call 'safetyism'?) Yes, that's basically what I mean (although it's of course a complicated social and psychological phenomena that can't be fully summarized in sentence).
wtfwhateven 3y ago
"the safety benefits are not worth the costs" is a determination government agencies already make with many things[1]. [1]: https://www.youtube.com/watch?v=xQbaVdge7kU https://www.youtube.com/watch?v=xQbaVdge7kU
- jessriedel 3y ago
  I'm very familiar with the idea discussed in that video, including the fact that it's in use by agencies like the NHTSA and the FAA. The problem is not that no one knows about these things, the problem is that (1) they can only be applied thoughtfully insofar as it's not part of public discussion (most people don't pay attention to NHTSA decisions) and (2) it's generally only applied when safety-at-any-cost is completely untenable. By (2), I mean the fact that in certain regulatory sectors you really can spend (or force others to spend) arbitrary amounts of money. Since there will be backlash against, that, regulators are forced to come up with a principle. But in many cases, the feasible restrictions can feasibly be dialed up to the max (e.g., complete banning of some product) and no one had to do a CBA.

ChuckMcM 3y ago

And yet the FAA makes that argument explicitly in the article.

jessriedel 3y ago
That's my point. The FAA makes the argument, and Eli Dourado should reject the framing (but doesn't). There are no flying cars in a future where the main arguments for reduced regulation is increasing safety. It doesn't take you remotely far enough.
- ChuckMcM 3y ago
  Fair enough, but that argument was not "for all types of aircraft" right? Are you arguing for the FAA to remove all safety requirements that have a cost? How do you evaluate that cost? When I read it I saw the FAA describing a methodology and argument that counters the "safety at any cost" argument (is that what you call 'safetyism'?) I read the article to say, "Sometimes our rules do not increase safety, sometimes our rules result in people choosing a less safe option because there isn't safer choice, and sometimes a new thing we haven't seen can be safer than what existing practice would expect." And MOSAIC was a response to addressing those observations in order to foster a more innovative and safer overall aviation industry. When I read your original point I mistakenly thought you were criticizing the FAA for lowering the cost of safety regulation when doing so increased safety. So that is how I got confused by your comment.
  - jessriedel 3y ago
    (No worries about the miscommunication.) > Are you arguing for the FAA to remove all safety requirements that have a cost? No. > How do you evaluate that cost? Lots to say here, but let me lead with a general principle: If you can measure the benefits of a regulation but cannot measure the costs, you should err on the side of less regulation. The original sin of cost-benefit analysis is "I can't easily measure contribution Y, so let's assume it rounds to zero". Here's one possible mechanism: Instead of banning an activity, either (a) put a tax on it equal to (say) 50% of dollar value of the safety harm (using, e.g., statistical value of life) or (b) require the consumer to pay for insurance that pays out a similar amount in case of accident. Then, even if we think the consumer is completely ignorant of the safety costs, they get to weigh the benefits to them of the activity against the financial costs from the tax/insurance, as determined by the regulators. Remember, the general intellectual justification of safety regulation is "We, the expert regulators, know the safety risks, but the consumer doesn't have enough information. Also, by typical paternalism arguments, it would be insufficient to simply create a Would-have-banned label ( https://mason.gmu.edu/~rhanson/wouldhavebanned.html https://mason.gmu.edu/~rhanson/wouldhavebanned.html ). It's very hard for us to estimate the cost of the regulation to the consumer, but based on eyeballing it, it seems smaller than the safety benefits." By taxing the activity at the level the regulators think, in their expertise, reflects the safety cost of the activity, they can let the consumer (who knows the benefits of the activity better) to decide whether the costs outweigh the benefits. Now, maybe if you implement that mechanism in the real world it turns out to be good or it turns out to be bad. Complicated empirical question. But the over-arching point is that no one even brings things like that up as possibilities. Which is evidence that the alleged justifications for the rules are not the real motivations. (Needless to say, the above suggestion is not new or invented by me; the ideas are just lying out there.) > When I read it I saw the FAA describing a methodology and argument that counters the "safety at any cost" argument It doesn't! It accepts the "safety at any cost" argument and then argues that, counter-intuitively, increased regulation reduces safety in this particular case. It's of course true that if a safety regulation both has costs and reduces safety then it should be removed -- it has no redeeming features. But we should be cognizant that this is an unusually easy situation. > the "safety at any cost" argument (is that what you call 'safetyism'?) Yes, that's basically what I mean (although it's of course a complicated social and psychological phenomena that can't be fully summarized in sentence).

wtfwhateven 3y ago

"the safety benefits are not worth the costs" is a determination government agencies already make with many things[1]. [1]: https://www.youtube.com/watch?v=xQbaVdge7kU https://www.youtube.com/watch?v=xQbaVdge7kU

jessriedel 3y ago
I'm very familiar with the idea discussed in that video, including the fact that it's in use by agencies like the NHTSA and the FAA. The problem is not that no one knows about these things, the problem is that (1) they can only be applied thoughtfully insofar as it's not part of public discussion (most people don't pay attention to NHTSA decisions) and (2) it's generally only applied when safety-at-any-cost is completely untenable. By (2), I mean the fact that in certain regulatory sectors you really can spend (or force others to spend) arbitrary amounts of money. Since there will be backlash against, that, regulators are forced to come up with a principle. But in many cases, the feasible restrictions can feasibly be dialed up to the max (e.g., complete banning of some product) and no one had to do a CBA.