Can you describe to me a situation in which reaching pareto efficiency would, in your opinion, not be the optimal state for the welfare of the aggregate? I'm trying to imagine a situation where resources are optimally distributed where that would be suboptimal for the welfare of everyone.

edit: I'm new to the concept of pareto efficiency, so hopefully this reads more as a sincere question rather than a challenge to your opinion

Sure.

Imagine a situation with two people and a state-like actor.

One person has all of the food in the entire world minus one loaf of bread. The other person has one loaf of bread. After they eat that loaf, they will probably die of starvation in the next 30 days.

This is a Pareto efficient distribution.

If the state-like actor intervened and re-distributed some of the food to the other person (with coercion), I would argue that aggregate welfare would be improved.

That's not a pareto efficient distribution. The efficient option would be for the person with bread to sell bread to the person without bread in exchange for future labor.

Pareto efficiency is defined to be an economic state where resources cannot be reallocated to make one individual better off without making at least one individual worse off. If a redistribution can make things better, it is not an efficient state.

> The efficient option would be for the person with bread to sell bread to the person without bread in exchange for future labor.

Okay, with work, let's say they are given the bare minimum to survive and they will get their work done (no rational reason to offer a higher wage). It would still be welfare enhancing to re-distribute.

> If a redistribution can make things better, it is not an efficient state.

This is just straight-up false, no other comment. Not what Pareto efficiency. The redistribution makes the person taken from worse off.

If they both have the bare minimum to survive, redistributing from one to the other puts the first below the minimum. If someone has a surplus, the efficient thing is to trade it. If they don't, redistribution hurts them as much as it helps another. There is no free lunch.

Again, that is the definition of pareto efficiency.

> If someone has a surplus, the efficient thing is to trade it.

With only two people, the most efficient thing to do is to trade it for the bare minimum the other person needs to survive. Beyond that is not Pareto efficient.

Yes, you give someone the bare minimum they need to survive. Then they want more. Now you're back to the situation where you have a surplus and they're willing to trade, the efficient thing is to trade your surplus. Continue repeating until one of you dies of old age, and the remaining survivor has all the bread.

So long as you have surplus, and that surplus has value, it makes sense to trade it, rather than have it sit idle.

Yes, someone living their entire life on subsistence amounts of bread while the other person has a great 3 meals a day is not a welfare optima.

I don't think you're understanding this. As long as person A has more bread than they're going to eat, and person B still wants more bread, it is still optimal to trade. A well fed individual can do more labor than a minimally fed person who can in turn do more labor than a dead person. As person B's quality of life improves, they are in an ever better position to trade. The bread is worthless to person A, it doesn't matter how little person B gives for it, it is still optimal.

It does not become Pareto efficient until person A is worse off by trading bread, by which point they no longer have enough to completely satisfy their needs.

This is a straw-man description of the ethics of Pareto efficiency. Even as such, if the second person has nothing the first person desires, what obligation does the first have to share their bread? Why can't the second person make bread or an alternative? Are they unable to, and if so, are they strictly a burden on the first person?

> what obligation does the first have to share their bread

None. However, part of the role of government (I would likely say it's entire role) is to make society as a whole better. It does this by imposing rules (obligations) on members of that society (taxes, a force to enforce the concept of private property, etc).

In the toy case described above, it is likely to the benefit of society to have both members survive and one member be unhappy; rather than have one member survive, but be happier. Hence, the government enforces an obligation on the first person to share their food.

Yes, you are coming from a very objectivist/libertarian perspective that I reject.

It's not a straw-man description of Pareto efficiency, it's a toy example designed to answer a very explicit question by the GP. What I described was a Pareto efficient distribution, and not the most welfare enhancing one.

What if neither person can make bread and the other just happened to inherit it all from their parents? Who is burdening who?

And that's before we get into the real-world ethics of whether we can honestly claim that contemporary wealth distributions are really derived from consensual interaction. Why should the descendents of nobility be entitled to wealth deriving from what are effectively violent takings?

I don't know - the whole deontological libertarian project in this sense seems like it has gaping holes.

I share the objectivist standpoint that, as long as wealth is accrued through fair transactions, having occurred without coercion (including monopoly influence) and because parties provided equitable value to each other, it (and its distribution to heirs over time) is ethically acceptable.

So then you'll agree that in a society where people without jobs are precluded from survival, labor is (in effect) coercive. Thus, any wealth accrued by arbitraging the labor of others was unethically gained and fair game for redistribution up to the point where joblessness is no longer an existential threat?

The fact that you need something to survive doesn't mean that every transaction providing it is coerced. Do you consider it coercion when you shop at the grocery store?

That makes sense, I guess there's subjectivity involved as to what that would look like as I imagine the death of one of those people would be an unwanted consequence to the person with all the food and as such the first person would be better off giving food to the starving person.

Your point makes sense to me now though, so thanks

> unwanted consequence to the person with all the food and as such the first person would be better off giving food to the starving person.

Doesn't really play out that way in actual real life (people still starve every day and both you and i probably have the resources to stop at least one of those starvation deaths), so didn't think of including this altruism in my toy model

Respectfully, I don't think we can say axiomatically that welfare is enhanced in the aggregate by Uber stakeholders ({drivers|employees|shareholders}) sacrificing in favor of the disabled because the Nth degree ethical consequences are unknowable.

My claim is that there is nothing about Pareto efficiency that is ethically preferable to a situation in which more people are better off.

Arguing that this policy would render some consensual exchanges illicit is irrelevant, because I don't think the most ethical outcome is axiomatically realized by solely consensual exchange without any coercion.