Hacker News Reader

Some mistakes can only happen once. Money doesn't fix that class of mistake.

One of the most annoying cliche arguments in American political discourse is that money is the universal 100% effective solution to all problems.

We just had a nuke plant melt down? That's going to just cost a lot of money! We want to provide universal top quality free on-demand healthcare for everyone? Just quadruple spending and change absolutely nothing else! We want to colonize mars in 10 years. That will just be a few trillion. No problem! Nine women working together to have a baby in 1 month? Just a few billion in government R&D bucks and we'll have it solved in a jiffy!

Argh, that's a frustrating fallacy, often posited with sincere intentions but often disconnected from practical implementation considerations.

In general terms, effectiveness and affordability of budget outlay should be the primary concerns. Instead, budget is typically orthogonal to effectiveness of quality of problem characterization, planning, procurement, execution and follow-up. Oddly, better results can sometimes be paradoxically realized by reducing budgets where there are capital-inefficient organizations by teaching resourcefulness... many bad habits emerge when people have too much cash available to spend frivolously. There are many instances where greater outlays are cheaper TCO-wise in the long run, so the "penny wise, pound wise" mantra tends to be a good guideline.

Also, in the US, defense spending, education and social security are sacred cows for various reasons.

Question military-industrial complex spending and be prepared for accusations of unpatriotism and treason for "not supporting the troops." These are completely specious but used to guard the hen-house. Robert Gates was the most recent, effective reformer of DoD, but the honeymoon didn't last long.

Education is seen as important human right however NCBLA has foisted short-term pseudo-results upon teachers. So the problem is still seen as not enough money, usually because administration bureaucracy typically consumes a disproportionate fraction of total budget (or in extreme cases, embezzles cash outright).

Social security is a tough one because many people have no direct retirement savings, and no one humane wants to go the way of Korea, where the elderly are de-facto shunned and have nothing.

Money is all we have left. We tried "sanctity" - more or less religious purity. We tried "honor" - ( bring back dueling? ). Duty, country, sacrifice - all pretty much deprecated.

I don't have an answer here - it's the stuff of Adam Curtis films. He's hardly perfect, but has an interesting way of constructing frameworks to look at these things.

If we ever need any reminders:

https://en.wikipedia.org/wiki/Hinkley_groundwater_contaminat...

http://www.damninteresting.com/undark-and-the-radium-girls/

http://www.southernstudies.org/2011/10/koch-industries-plant...

http://www.democracynow.org/2004/12/6/yes_men_hoax_on_bbc_re...

https://en.wikipedia.org/wiki/Chernobyl_disaster

https://en.wikipedia.org/wiki/Exxon_Valdez_oil_spill

and plenty more

The moral of the story is libertarian idealism (sans facism/authoritarianism) is simply the mirror opposite utopian extremism of communism... ever-refined balance is needed, which tends towards democratically-elected, semi-socialism: "If men were angels, no government would be necessary." - James Madison

What about something on the scale of Bhopal (https://en.wikipedia.org/wiki/Bhopal_disaster)?

See the 4th link.

I remember a story from my Operations Research class ( most likely apocryphal ) that Stalin, riding a coal train to the front, watched as another coal train passed them and said "have whoever schedules the trains shot."

The oil industry isn't even a close second to agriculture in terms of risk to freshwater resources in the American West ( where the story takes place ). After a while as you read about our West, you realize that people may actually not be all that much supposed to live there at all, at least in parts of it. Trying to elevate freshwater to "irreplaceable" ignores almost all the behavior that actually surrounds it.

No, really - there's a book - "Cadillac Desert" which provides much context for this.

The fact that regulation seems always end in capture tells me something. For one, it's very hard to do. Think about how hard it is to assemble 100 people who truly understand something that is very complicated deeply. Something that takes dedicating your life to. Now think about how you'd assemble say, 200 of them. It's be some exponent or power harder, wouldn't it?

Criminal penalties may sound like a good idea ( to some; not to me ) but I don't really have a way to take such statements at face value. I'm not prepared to take what is generally considered legitimate activity and round up certain classes of mistakes to "go to jail."

Investing groundwater with "rights" and the ability to suffer "criminal torts" ( regardless of the horrors that may be inflicted on it ) seems a bridge too far. Even if we sum the rights of all the people who depend on it, it's difficult. After all, people have had to leave places before.

There could ( and perhaps should ) be some authority that must approve such activity in certain places. In the example, there should be no messing with formations that are that shallow. But that approval process sounds easier than perhaps it is.

>No. The whole point is that these "risky behaviors" are quite capable of inflicting damages (such as the permanent contamination of water tables in certain areas) that can't possibly be compensated by the any conceivable "insurance payout"

Those are the extreme exception, not the rule. That might be true of nuclear operations, but everything else can be fixed at some cost, with maybe ~$1 trillion for tail end risks, easily coverable by reinsurance networks.

Environmental damage is bad, but it's not infinitely bad. If the costs were anything like you suggest, that would justify vigorous bans of most industrial operations, not some mere regulatory enforcement as could be accomplished even with jail time.

Those are the extreme exception, not the rule.

Right, it's not the average case we're discussing here. But I think we can both agree that wrecking a water table in a geographic area of significant size (for example) would qualify as "extreme."

Everything else can be fixed at some cost, with maybe ~$1 trillion for tail end risks, easily coverable by reinsurance networks.

No - plenty of things in life cannot be fixed at any price.

Environmental damage is bad, but it's not infinitely bad.

Between "bad" and "infinitely bad" (the latter being obviously a red herring) there's another category which would seem to apply for the scale of mishap here: "unacceptably bad."

If the costs were anything like you suggest, that would justify vigorous bans of most industrial operations, not some mere regulatory enforcement as could be accomplished even with jail time.

Again, you're arguing by referring to extremes: "Yeah, drunk driving carries risks, but so does sober driving. If we really want to eliminate risk of death and dismemberment in horrifying accidents, then, we should ban all forms of driving."

No -- I'm just saying we should ban drunk driving. Which -- the more we get into the basic research behind these activities -- (certain classes of) fracking and other forms of large-scale resource extraction are increasingly looking to be aptly analogous to.

>Right, it's not the average case we're discussing here. But I think we can both agree that wrecking a water table in a geographic area of significant size (for example) would qualify as "extreme."

No, we can't. In the context of the specific point you were making, extreme means "no feasible amount of money could undo the damage". That's not true of damage to a water table, which can be remediated with a feasible amount of money.

>... I'm just saying we should ban drunk driving.

No, the point here is that, if drunk driving's effects were as uncountably bad as you claim the potential environmental damage is, we should ban driving, because the effect would wipe out all the benefit of driving, since the insurance cost would make it infeasible. But that's absurd, so we must conclude that your estimates of the harm size are as well.

Remember, you're the one arguing that the extreme cases dominate the risk calculation (to the point of uninsurability), not me!

No, we can't [agree about large-scale water table damage being remediable with feasible amounts of money].

Well, we disagree then.

As to the other stuff -- it seems you're doubling down on extremification techniques, attributing words ("uncountably bad") to me that aren't mine. All I can say is I don't find that line of reasoning helpful, and I don't know what to say to it.

There are no "extremification" techniques; it's just that you made a very specific claim about the nature of environmental risks, that such harms are typically (at their worst) so bad as to be uninsurable.

My point is that this claim cuts both ways -- in order for the environmental damage of industrial disasters to be that bad (above the ~$1 trillion that a reinsurance network could cover), it would also have to be so bad as to justify banning most classes of industrial activity -- not allowing them even with regulation. Since you probably don't believe that, then you're probably overstating the mangitude of this tail risk, and dismissing its applicability to your own alternative.

(Remember, regulation doesn't turn off the extreme events; it's just one way to reduce their likelihood.)

Which technique do you think that argument is unhelpfully using?

It's just that you made a very specific claim about the nature of environmental risks, that such harms are typically (at their worst) so bad as to be uninsurable.

"typically (at their worst)"?

I'm not even sure what you mean by that. (I have some guesses at what you might be getting at, but they also point to things I didn't say, either).

In order for the environmental damage of industrial disasters to be that bad (above the ~$1 trillion that a reinsurance network could cover), it would also have to be so bad as to justify banning most classes of industrial activity.

(1) I don't agree (nor do I think most people would agree) that all environmental or human disasters have measurable impact (let alone have a dollar value -- even a very high one -- which would suffice to repair or compensate for the loss the incur).

(2) The second half of your statement does not follow from the first.

Which technique do you think that argument is unhelpfully using?

https://en.wikipedia.org/wiki/Reductio_ad_absurdum

>"typically (at their worst)"?

>I'm not even sure what you mean by that.

Then review the original usage, which should have made it clear -- I was distinguishing operations where (I claim) the worst case is insurable from the cases where it isn't. I remarked that the worst case is rarely uninsurable, and gave the example of nuclear operations (which means nuclear power and nuclear weapons).

So, typically (across the set of environment-risking things), the worst case (within one of those things) is not beyond the reach of insurance.

Sorry if that whole "typical extreme" thing threw you for a loop, but I thought it was clear from context. Let me know if you know of a better way to communicate this concept. You seem to be deeply familiar with these kinds of considerations, so I bet you have one.

>(1) I don't agree (nor do I think most people would agree) that all environmental or human disasters have measurable impact (let alone have a dollar value -- even a very high one -- which would suffice to repair or compensate for the loss the incur).

Well, then you've inserted infinities in your model, which fundamentally breaks things and renders further debate pointless. The moment you say "this outcome is so bad that it dominates all other outcomes", then even small probabilities of it are unacceptable -- the Pascal's Wager dynamic. But that equally cuts against regulatory approaches, and it's not a reason to favor top-down regulation vs the insurance system advocated in the thread.

Generally speaking, the moment you refuse to quantify the damage, it's no longer in the reach of reasonable debate. "This thing is infinitely valuable" makes for nice platitudes, but not sound policy.

>Reductio ad absurdum

It's valid to point out how someone's claim implies an absurdity they disagree with :-p Certainly, you can dispute that the implication follows, but I'm not sure where you're coming from with a categorical dismissal.

So, -typically- (across the set of environment-risking things), the -worst case- (within one of those things) is not beyond the reach of insurance.

Right -- I think I've been making it pretty clear, though, that I was referring to high-risk industrial activities, not to average case ones.

Generally speaking, the moment you refuse to quantify the damage, it's no longer in the reach of reasonable debate. "This thing is infinitely valuable" makes for nice platitudes, but not sound policy.

I disagree; I find that qualitative reasoning can often be quite useful. Especially to avoid the pitfalls of purely quantitative reasons, which can often lead one to either-or arguments (such as you are invoking above), reduction to extremes, or other cul-de-sacs of the mind.

In particular, you'll note that I've never used the phrase "infinitely valuable" (or "infinitely bad"), and have in fact specifically eschewed it, despite your repeated attempt to assist the contrary.

All I'm referring to, basically, is the fact that some forms of damage are basically irreparable -- which is quite different from saying "the thing is infinitely valuable" (or that wrecking it is "infinitely bad"). In other words, it's a distinction of kind, rather than simply degree. And all I'm saying, policy-wise, is that industrial activities which carry a high enough risk of large-scale, irreparable environmental damage are of a different kind, such that it is no longer reasonable to believe that insurance requirements (or any threat of civil penalties against the corporate entities engaging in them) will provide adequate protection; instead they should banned outright.

Like, you know -- drunk driving.

I don't see this as an unreasonable position, not does is it (as you would seem to suggest) undermine our axiom systems to such a degree that everything becomes unknowable and all debate becomes pointless. Rather, it's just common sense, qualitative reasoning -- that's all.

The kind of damages that can't possibly be compensated by any conceivable insurance payout are also not helped by some people being in jail for life, so this isn't the solution either.

But it may be the best we can do.

Otherwise, it'd be like telling drunk drivers "Hey it's your call, bro! You just need to wear a seatbelt and carry lots of insurance, 'k?"

In our current legal and regulatory framework, companies have to seek licenses from the government to carry out these activities. So it seems to me you are advocating handing out 'serious jail time' to people who carried out legitimate business activities, regulated by the government and for which they obtained legal licenses.

Or am I missing something?

The fact is environmental impact should be taken into account properly at the licensing stage. That's where the accountability falls.

it seems to me you are advocating handing out 'serious jail time' to people who carried out legitimate business activities, regulated by the government and for which they obtained legal licenses.

No, I'm saying that high-risk industrial activities shouldn't merely be "regulated"; they should be banned out right.

It's when you violate the ban that you get handed a jail sentence.

>And BTW by "jail time" I mean serious jail time, as in multiple decades of jail time, and/or life sentences if necessary.

Have severe jail sentences rid the world of illicit drugs users, murderers, cutpurses, burglars, vandals, fraudsters, rapists or anything else?

If the chance of being caught and going to jail is lower than the expected gains of not getting caught, criminals gonna criminal. How do you aim to stop that with a threat? Throwing someone in jail is "too little, too late".

> Have severe jail sentences rid the world of illicit drugs users, murderers, cutpurses, burglars, vandals, fraudsters, rapists or anything else?

No, but you have to put things into context. Prison is a lot worse for a 50+ year old with a few million in assets than someone who's only other alternative is working at a fast food restaurant for minimum wage. When the choice is deal drugs or simply not make enough to survive there really isn't much of a choice.

Good government prevention detects such attacks on the allmende before they happen. If you are going to make a paint factory, and you have no contracts, and no money paid, to handle your waste, you are a crook. If you have a paint-production-recyling factory and you have no means to safely handle the remains, you are a crook. This chains can be checked, the quantities handled can be booked. Its a lot of bureaucracy - but its worth it. And i say that, having worked in one of those "regulated" factories and having a inspector being a pain in the ass. Its really, really necessary. Just saying "Nothing can fix this." is easy here, but i dare you say that to the face of a mother or father who just lost there child to cancer. So not even prison- just deny the freedom to be a parasite- to public goods.