Hacker News Reader

Comment by jessriedel | Hacker News Reader

If no one gets injured on a $500M project, safety standards are almost certainly too stringent.

I feel like this comment is callous, but unfairly downvoted.

All projects contain risk. Getting out of bed and taking a shower involves risk, you can slip and fall. The larger the project, the larger the sample size and the more statistical considerations for risk should be managed.

Modern western society has greatly lost all tolerance of risk. One difference that struck me for risk culture, compared to the past, is old footage of the 1955 Le Mans disaster: after the catastrophic crash, bystanders themselves helped carry bodies away from the scene, and clean up wreckage. This calmness and social damage tolerance in the process of a disaster is very different from a modern "evacuate and let the professionals handle it" attitude, and I think this is an example of a clear shift in social tolerance of risk.

I wonder if the Apollo project would have worked today, instead of in the 1960s. At the scale of an infrastructure project of half a billion dollars, we're not talking about something so big that zero injuries safety standards are bad. However, as you get larger projects, it is statistically impossible to keep risk at 0; I'm not sure how willing modern society is to accept that injuries on large projects is inevitable.

Elsewhere in this thread, user JumpCrisscross comments about comparing the cost overrun to the second bid. Following up that comment, the top response is something along the lines of "nobody was injured, so showering the contractor with money is ok". This attitude is prevalent in modern society, and it doesn't really get questioned. Let's pose a hypothetical scenario: If a person has their leg broken, let's say a platform caused someone to fall and hurt their back. Would that amount of damage be worth a $250million additional cost? Think of the amount of social good you can do with $250million. At what level does society overvalue the reduction of risk? This is an important question to ask.

Let's not even talk about minor injuries. Let's say 1 person died. If you walk up to a random person, tell them that they have the opportunity to die; BUT their immediate family would receive $125 million dollars, and $125million would be donated to the charity of their choice. I suspect a very large percentage of people would actually be willing to take that offer.

And yet, giving an extra $250million to a contractor to build something is considered "worth it" to the general public if no injuries were reported. At which point does society overvalue reducing risk?

So what value would you put on someone being killed or maimed to save some money? Because that is exactly the calculation you are proposing should happen.

A quite a bit less than $250million, which is how much is being paid.

In fact, people would even volunteer to die for "just" $100million. Just see this reddit thread: https://www.reddit.com/r/AskReddit/comments/4jqfka/what_woul...

This is EXACTLY what I mean by "risk management for large projects". For a small $10k project? A small percentage cost decrease for increased risk of death isn't worth it. For a multi-billion dollar project? All of the sudden, the incentives change drastically. How many people would be willing to work on a project that is a tiny amount more risky, but knowing that their families will be handsomely compensated in case anything happens? If you told me that my family will receive $100mil guaranteed if I die/get maimed, there's very little I'm not willing to do.

Also, to note: $100,000,000 is enough to save around 30,000 children from an excruciating death by malaria.

TL;DR 1 average human life generally isn't worth throwing away $250million. I will literally volunteer my own life on this fact.

But how do you know that it is a simple trade off between money and safety standards? It's been noted on HN that the US seems particularly bad at large infrastructure projects compared to other developed countries and I'm struggling to believe that's because the US has wildly better safety standards.

I don't. That's moving the goalposts.

Who cares if Oroville Dam is in the USA or not. The same thing would still apply if it was in the UK. I'm merely saying the modern public should reconsider spending $250million extra, instead of applauding it simply because no injuries happened.

I'm going to make a guess here and assume that you are not working construction. And if that's so I would highly recommend you join a construction crew for a couple of years and then come back here with your idiotic approach to workplace safety. It's all about preventable accidents, and your assumption that we could save 250M on a project by compromising workplace safety is both wrong and the beginning of a slippery slope that ends with maximum savings and no workplace safety at all. After all, it's not the life of you and your family that is at risk here.

Fuck those construction workers, fuck those miners and who cares about factory workers. After all they're disposable. /s

Workplace safety is there for a reason: to stop white collar criminals from putting a price on human lives and then to optimize for the lowest they can get away with.

I feel you're missing the point, I can't tell if it's deliberate.

Suppose you wanted to reduce risks for computer programmers - everyone gets a personal doctor who sits with them to monitor their vitals, state of the art monitoring device, daily blood tests, screenings for all sorts of diseases. Each person gets a fitness coach, dietician, sleep coach, whatever. Programmers with injuries get oxygen therapy and physios on hand. Basically you treat them like world class racehorses (!). Whatever, hopefully you see the point.

Would that be overkill? No? Well then you don't think programmers matter at all and want to see them all die in industrial accidents. /s

If we can accept at any point that we should, maybe not buy the most expensive chair/table/keyboard/screens, even if there's a chance of injury (eyesight, cardio problems from 20 years of sitting, cut from a sharp edge on a filling cabinet, etc.) then you're agreeing with the general concept of the OP IMO.

Now, in public procurement, should we pay $1000 per seat for a currently $100 app to ensure no programmer ever gets a thrombosis/carpal problem/etc.?

So you are saying you believe that if a project completes with no deaths or injuries then too much money was spent on it?

I'm saying that as size of project (lim -> $infinity), (probability death or injury) approaches 1.

I agreed with some of your more abstract posts on this topic, but I really can't buy into that assumption that just being a little less fearful would reduce costs by half. You wave that second bid around as if they had actually done the same work, at the lower price, with a known quantity of bloodshed. But that is only a hypothetical, That second bid might have had even bigger cost overruns, we don't know. The winning bid might actually have terrible safety standards while being lucky, we don't know. There is no simple, isolated safety/cost tradeoff that you can dial in at will and get the expected results. On the contrary, a contractor that is sloppy with safety will often also be sloppy with other things, causing costs to increase, not to sink.

I agree, I'm only arguing the hypothetical.

I would highly doubt that $250mil went into slower progress for the sake of safety. Definitely, there were other costs.

You're supposed to only consider the abstract idea I presented. That's why I changed from $250mil numbers to $100mil halfway through, although I didn't make it fully clear to divorce the concrete numbers from the abstract cultural ideas.

> A quite a bit less than $250million, which is how much is being paid.

That idea was made up in this thread, it doesn't reflect reality.

The cost overrun was explained in the article as due to only scoping out 30% of the project as part of making the bid, and as such underestimating how much concrete they would need:

> Jeff Petersen, project manager for Kiewit, said that once construction workers got on the site they discovered they had to dig much deeper to get down to bedrock than they had expected. That meant they are having to pour almost twice as much concrete this year — 870,000 cubic yards — as the company and the Department of Water Resources had planned.

To be willing to throw your own life away for $250M is one thing, but to impose that choice on others, so that the government or some large contractor company gets to save money, smells of an utilitarian view of society and is frankly immoral.

And that is the exact same calculation the local transportation agency makes every time their is a car accident with fatalities on a major undivided highway / road. Roads become known as dangerous (eg Blood Alley, officially the 126 is Southern California), and clearly the govt could step in and install some Jersey barriers. Yet they don't, because implicitly the cost of X lives per year is not worth it to society.

But I hope whoever does the calculations for you doesn't decide to sell safety barriers from roads because nobody died on them.

I was objecting to the idea that having nobody dying on a project being a bad thing because it implies overspending on safety.

I love it when deskjockeys start putting a price on other people's lives.

Construction is dangerous, if you are willing to compromise on that then you'll see that you can do even better when you drop that price. And before you know it we're back in the 1930's again.

The simple fact is that there is a trade-off between the effective safety measures put in place, and the costs of doing so. And at some point you cross a line between reasonable and unreasonable expense. Dollars may be imaginary, but they trade for real resources, and those resources exchange for opportunity costs elsewhere, including lives that might be saved or improved by other routes.

I'll but the question back to you: how much are you willing to pay to save one additional life? $1 million? $1 billion? $1 trillion? $100 trillion? (That last is more than total gross world production.)

I've got my own considerable problems with numerous elements of evaluating costs and benefits in economics, many of which I've been trying to articulate for over three decades and am only just now starting to get a handle on. I would not argue, for example, that just straight earning potential is an adequate basis. I also don't think that various proposed "market bid" rates for how much individuals are willing to accept under very specific and limited conditions are particularly useful.

But even accounting for community, social, and disruption losses, somewhere between tens and hundreds of millions is likely a reasonable balance.

That's fine as long as the ones deciding on the balance have to take a large share of the risk of injury. Be the first to work in the same conditions you expose others to. Otherwise, you're just a hypocrite and a monster.

Your rhetoric doesn't match the world.

Part of society involves distributing decisionmaking and risks. And no, not always equally. I don't find the argument that those most adept at accurately and fairly assessing risks are necessarily those who can accomplish the tasks involved.

To put this in a different field: I'd not expect my airline pilot to be the best qualified accident investigator, nor systems engineer, nor aircraft fabricator. And I as passenger bear the consequence of their combined work.

Reality bekons, if you care to heed its call.

'No taxation without representation' is not based on the presumption that those being taxed are the best at determining the value of taxation, or even of electing their own representatives. Those who do not bear the cost of their own decisions are a priori unqualified to make those decisions as they are subject to adverse incentives.

Well, yeah. They can opt to not pay the "risk tax", by simply not going down that career path. Not every construction job is rebuilding a high risk dam spillway. Not every aerospace engineer job involves stepping into a 100% oxygen flammable Apollo rocket chamber.

The understanding is that 1. society should be allowed to consider increased risk, and 2. in the event something bad happens, the victims should be compensated.

Literally nothing I said stated about society and risk stated that dangerous jobs are only for low level construction workers. You're just using silly ad hominem ("desk jockey") and emotional appeals (poor vs rich!) without considering the point.

In fact, one of my examples of risk was the Apollo project, where the most notable 3 casualties (Apollo 1's Virgil Grissom, Edward White II, and Roger Chaffee) were highly paid astronauts with engineering backgrounds. I'm not discriminating against construction workers, I'm making a statement on acceptable risk in society in general.

Finding a balance on putting price on human life is simply a necessity. I have a friend who's a petrol engineer working dangerous jobs at a high salary, and it's a risk he's willing to take. However, the modern general public is culturally shifting to people like you who literally hear the words "possible risk" and the frontal cortex of the brain shuts down and start ignoring the costs.

You've put an interesting perspective on the safety/cost balance. I find it ironic that the replier is effectively demonstrating your point: people have an emotional attachment to safety, but not opportunity cost, regardless of the math. This is perhaps a form of "one death is a tragedy, ten thousand deaths are a statistic."

>I wonder if the Apollo project would have worked today, instead of in the 1960s.

George Mueller (who was head of NASA during the Apollo program) said that it would be impossible to repeat Apollo today without it being a classified program as Congress has passed so many laws and regulations in the meantime.

Source: https://dominiccummings.com/2017/06/12/the-unrecognised-simp...

I honestly think this is just reflection of how the value of a human life has changed in the last few decades. The thought of spending millions of dollars you don't have to save a newborn with a serious heart defect is praiseworthy. The thought of bankrupting yourself and your loved ones for a chance at another year or two on the earth is normal.

What I can't say is whether this is a good thing or a bad thing. There's a part of me that recoils at the thought of putting a price on a life. Then there's another part of me that recoils at the selfishness of bankrupting my wife should I get cancer. That same selfishness recoils at the thought of bankrupting myself should she get cancer.

I'm not looking forward to those decisions.

Risk tolerance has definitely gone down, but what’s the basis for assuming that the level of tolerance was right in the past, and today it’s too low?

Historically, the people investing in safety were not the people who were exposed to the risk. That created perverse incentives, where $1 in safety improvements that would save an average of $10 in human health or life would be rejected because the people who would pay the $1 would not be the ones paying the $10.

That’s still true, of course, but we at least attempt to make people do the right thing.

Regulators today aren’t afraid to assign a dollar value to a human life and reject safety improvements that don’t have a positive ROI. The standard value currently is something like $8 million.

I suspect that the parent wasn’t downvoted for callously making a good point. They were downvoted for baselessly assuming that the correct level of safety tolerance would result in injuries or deaths on this project, and doing so callously.

Considering that even 100 years ago, such a project would have figuratively (and sometimes literally) mixed a good bit of blood into the mortar, yeah.

Having worked a little bit of industrial construction in a past life, it's amazing anything gets done at all, if you do follow all the rules. And even then all it takes is one slipup to get somebody hurt.

In that case, what's the "correct" number of deaths that safety standards should target per $1M spent?

The statistical value of a human life in the US is ~$10M.

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

Serious injuries are some fraction of that, e.g., ~$1M for a limb. The correct tradeoff between safety and cost obviously depends on the marginal cost to increase safety, which is known to the project leaders but not us, but I'd be surprised if the optimal number wasn't between 5% and 25% of total cost. 5% of $500M is $25M, so for a project of this size I'd expect, as a lower bound, at least one death and a dozen serious injuries.

To be clear: by not making this trade off and by spending huge amounts of money to ensure such tiny injury rates, we are wasting resources that could be spent on saving lives at the rate of $10M/life. Indeed, this number is the key threshold by which the EPA decides which environmental interventions to fund, or the NHTSA decides which traffic safety measure to take.

Are you volunteering?

Reply on news.ycombinator.com