[coldtea]

>That trade off is deeply flawed. Future lives saved are impossible to quantify, while lives saved now very much are.

Actually both are "future lives" (we're not saving someone we already know is about to die) and both are as easy or hard to quantify.

We can quantify average cost of lives in a big building project given X regulations applied (based on past experience with similar projects), and we can also quantify the impact of a project in first (and sometimes second) order effects (e.g. "building this highway bypass means X less traffic and congestion in that populated area, and thus X man-years spared, plus X less pollution").

[zaphar]

They are in no way equivalent in relative ease to quantify. Safety Regulations exist to protect from "known" harms. It is the result of learning from our past and applying that knowledge to our present and future. But anything in the unknown harms category is fundamentally unquantifiable. We know there are unknown harms out there. But we don't know

1) What they are

2) When they can happen

3) How many they could affect

You can't make decisions that help you avoid harms in that category. But you can make decisions in the other category. There is an argument that failing to build is for example in a category of known harms now. But so are a whole host of harms that we encountered to get where we are now. The trick is to make it easier to build while not also engaging in the rest of the known harms out there.

[coldtea]

>But anything in the unknown harms category is fundamentally unquantifiable

The "unknown harms" exist in workplace deaths. In fact many of them come from such factors. It's not just things like "asbestos is bad to work with, you need a face mask and other protection" or "construction debris might fall, workers need to wear a hard hat" and such, but also things like materials breaking under unforseen circumstances, unforseen disasters like earthquakes and fires while on the job, and more...