4 ms·

There's also work where the overall utility generated might be much higher than the cost of the labor, but the utility generated specifically for the business o

There's also work where the overall utility generated might be much higher than the cost of the labor, but the utility generated specifically for the business or local government considering the project is lower than the cost of the labor. Planting trees might only be worth $5/hour for the local lumber company, but the existence value of a new forest to the population generally might be many times that. I suspect that the amount of potential work that fits into this category is enormous.

int_19h 9y ago
This is an orthogonal issue, though - we're just not good at capturing the true benefits (and the true costs, including externalities) for many things. Broadly speaking, though, I agree. This is one other case where we can do better - instead of constructing a "broken window" in the private sector and have the person repair it to keep them busy, directly utilize their labor for a socially productive task that may not be market-efficient. But I'm not sure that there's enough of that kind of useful work (i.e. the kind that actually makes a difference) to keep everyone busy. It also feels like it's mostly easy to automate.
- kwhitefoot 9y ago
  > But I'm not sure that there's enough of that kind of useful work Perhaps we should not worry about that until all the useful work has in fact been done.
  - int_19h 9y ago
    If we're retraining people to do it, we absolutely should, because we'd be spending significant resources for something that might prove a very short-sighted investment. It might be better to just invest that money into automation to do that same work instead.
    - dagw 9y ago
      I imagine that most jobs that have a 'value' of ~$5/hour don't require much training. Also we could implement this policy tomorrow if we wanted to while automation is still many years out (at best) for a lot of tasks, so the two policies could be run in parallel.

int_19h 9y ago

This is an orthogonal issue, though - we're just not good at capturing the true benefits (and the true costs, including externalities) for many things. Broadly speaking, though, I agree. This is one other case where we can do better - instead of constructing a "broken window" in the private sector and have the person repair it to keep them busy, directly utilize their labor for a socially productive task that may not be market-efficient. But I'm not sure that there's enough of that kind of useful work (i.e. the kind that actually makes a difference) to keep everyone busy. It also feels like it's mostly easy to automate.

kwhitefoot 9y ago
> But I'm not sure that there's enough of that kind of useful work Perhaps we should not worry about that until all the useful work has in fact been done.
- int_19h 9y ago
  If we're retraining people to do it, we absolutely should, because we'd be spending significant resources for something that might prove a very short-sighted investment. It might be better to just invest that money into automation to do that same work instead.
  - dagw 9y ago
    I imagine that most jobs that have a 'value' of ~$5/hour don't require much training. Also we could implement this policy tomorrow if we wanted to while automation is still many years out (at best) for a lot of tasks, so the two policies could be run in parallel.