3 thoughts on “2-3 Athena Aktipis. “It all comes out in the wash”: Risk pooling and need-based transfers in The Human Generosity Project.”
Ole Peters
Superb! Thank you, Athena.
I had to look up what you meant by debt-based transfers. So I googled and found this paper https://pubmed.ncbi.nlm.nih.gov/27445430/ and I think I understand now. Debt-based transfers are more restricted, in the sense that there can be situations where someone has enough to help but doesn’t because the agent requesting help has maxed out his credit line (and similar instances).
It looks like there’s a hierarchy of generosity. The system you’re studying is very similar to our work on reallocating geometric Brownian motion, or on cooperation.
Looking at all of these different studies
* our cases (in the EE group) are unconditional cooperation. No one asks, it’s not need based, not debt based, people just pool and share. The result is high growth and low risk of ruin.
* The next level is need-based: no matter what you’ve done before, if you need help, we’ll help. This has a slightly higher risk of ruin and probably a lower growth rate (something to look into?).
* Then comes debt-based: again a little higher ruin probability and lower growth rate (something to check).
* Finally, there’s no cooperation at all: everyone for himself, and this results in even higher ruin probability and lower growth.
These benefits of cooperation will be visible whenever ergodicity is broken in the way it is in the cases our groups have studied. We’ve mostly focused on systems where ergodicity is broken because growth is non-linear (that’s also true for you, actually), and you have mostly focused on systems where ergodicity is broken because an absorbing ruin state exists.
It’s really great to see that we’ve arrived at such similar models for studying cooperation. It’s very easy to see them as different versions, or extensions in different directions, of a basic common process. Lots to discuss!
The need-based transfer system that you described made me think of Rawls’ Maxi-Min principle of social justice. It implies that maximising the social welfare of a group of individuals amounts to maximising the smallest individual wealth (or utility). Hence, according to Rawls, redistribution of wealth is only justified if it helps the poorest in the group. Your research suggests that what is just (according to Rawls) also appears to maximise survival.
There has been some discussion on whether the Maxi-Min principle can be applied to inter-generational wealth transfers (through savings) and I am wondering if your research offers any insights regarding this.
I really appreciate the distinction you highlight between needs-based and debt-based cooperation and it is really interesting that both systems operate in the Masai society at the same time.
A few questions:
Are there any societies you have studied that use the “pool and share everything equally” form of cooperation that the Farmer’s Fable models and that Ole mentions above?
Do you see a meaningful difference between the needs-based cooperation and the simpler “pool and share everything equally” form. It strikes me that the needs-based might actually be simpler in practice as it would avoid any central coordination of accounting and redistributing, but achieve more-or-less the same benefits of moving resources to where they are needed.
Is there a functional difference between the debt-based system and simply exchanging goods for a commonly agreed currency.
Superb! Thank you, Athena.
I had to look up what you meant by debt-based transfers. So I googled and found this paper https://pubmed.ncbi.nlm.nih.gov/27445430/ and I think I understand now. Debt-based transfers are more restricted, in the sense that there can be situations where someone has enough to help but doesn’t because the agent requesting help has maxed out his credit line (and similar instances).
It looks like there’s a hierarchy of generosity. The system you’re studying is very similar to our work on reallocating geometric Brownian motion, or on cooperation.
Looking at all of these different studies
* our cases (in the EE group) are unconditional cooperation. No one asks, it’s not need based, not debt based, people just pool and share. The result is high growth and low risk of ruin.
* The next level is need-based: no matter what you’ve done before, if you need help, we’ll help. This has a slightly higher risk of ruin and probably a lower growth rate (something to look into?).
* Then comes debt-based: again a little higher ruin probability and lower growth rate (something to check).
* Finally, there’s no cooperation at all: everyone for himself, and this results in even higher ruin probability and lower growth.
These benefits of cooperation will be visible whenever ergodicity is broken in the way it is in the cases our groups have studied. We’ve mostly focused on systems where ergodicity is broken because growth is non-linear (that’s also true for you, actually), and you have mostly focused on systems where ergodicity is broken because an absorbing ruin state exists.
It’s really great to see that we’ve arrived at such similar models for studying cooperation. It’s very easy to see them as different versions, or extensions in different directions, of a basic common process. Lots to discuss!
Thank you for this really interesting talk!
The need-based transfer system that you described made me think of Rawls’ Maxi-Min principle of social justice. It implies that maximising the social welfare of a group of individuals amounts to maximising the smallest individual wealth (or utility). Hence, according to Rawls, redistribution of wealth is only justified if it helps the poorest in the group. Your research suggests that what is just (according to Rawls) also appears to maximise survival.
There has been some discussion on whether the Maxi-Min principle can be applied to inter-generational wealth transfers (through savings) and I am wondering if your research offers any insights regarding this.
Great talk, Athena! Thank you!
I really appreciate the distinction you highlight between needs-based and debt-based cooperation and it is really interesting that both systems operate in the Masai society at the same time.
A few questions:
Are there any societies you have studied that use the “pool and share everything equally” form of cooperation that the Farmer’s Fable models and that Ole mentions above?
Do you see a meaningful difference between the needs-based cooperation and the simpler “pool and share everything equally” form. It strikes me that the needs-based might actually be simpler in practice as it would avoid any central coordination of accounting and redistributing, but achieve more-or-less the same benefits of moving resources to where they are needed.
Is there a functional difference between the debt-based system and simply exchanging goods for a commonly agreed currency.