Hi everyone, I’m writing this post as a relative newcomer to the community. I’m a professor at Arizona State University who works on cooperation across systems—I study topics including how humans share in small-scale societies and how cells in our body cooperate to make us function. I also work on building interdisciplinary and creative communities for working on problems that require large-scale information sharing, including grappling with catastrophic risks (e.g., the Zombie Apocalypse Medicine Meeting). I would like to contribute to this community and learn from you all as I continue working on these issues and building communities around grappling with these problems.
One of the topics that I study is how small-scale societies manage risk though sharing networks (called need-based transfer systems or risk pooling systems), and how we might be able to scale up these kinds of local sharing networks to deal with larger risks. I understand that the focus of EA is very much on larger-scale prosocial behavior rather than building local communities (I’m saying this as a newcomer to the community, so please correct me if I’m wrong), but I think there is the potential to use these need-based transfer principles as a sort of building block for creating multi-level risk management systems that are rooted in communities (whether those communities are in suburban Arizona or rural Africa). These kinds of networks can provide a buffer against many different kinds of risks (e.g., drought, disease, extreme weather events). I wrote more about how these principles work in a blog post based on a paper my colleague and I published in Nature Human Behavior. We found that successful need-based transfer (risk pooling) systems followed these seven key design principles:
Insures against the unexpected. Participants should agree that the pool is for needs that arise unpredictably, not for routine, predictable needs
There is no obligation to repay. Giving to those in need should not create an obligation for them to repay
Givers are not at risk. Participants should not be not expected to help others until they have taken care of their own needs
Consensus about need. Participants should have a consensus about what constitutes need
Wealth should be visible. Resources should be either naturally visible or made visible to reduce cheating
Partner choice should be allowed. Individuals should be able to decide which partners to accept
Network scale should cover risk. The scale of the network should be large enough to cover the scale of risks
Need-based transfer networks also provide valuable opportunities for sharing information. People in need-based transfer networks share information about the risk landscapes that the occupy, both explicitly and implicitly. When they make requests for help or provide help to one another this indirectly shares information about the risk landscape because requests only happen when people encounter challenges.
Also, the simple existence of these networks creates the opportunity for common knowledge around the risks that communities face, which facilitates coordination and cooperation around managing risks as a community. This is part of why I think that it can make a lot of sense to invest in building communities around such principles, whether we are talking about cultivating local communities or building intellectual and scholarly communities to better understand and manage large-scale risks.
Need-based helping networks are also often a foundation for of social, economic and cultural life in small-scale communities, providing other valuable functions that it can be hard to quantify. It is likely that they contribute to improved quality of life, lower stress and better health outcomes overall, given that they provide a sort of general support network (and these kinds of benefits as associated with being embedded in support networks).
I realize that these kinds of need-based sharing networks might not necessarily be able to do that much to mitigate risks that many people in the EA community are concerned with (e.g., AI risk), but there may be ways to leverage other things we know about how small-scale societies manage risk to do a better job managing the risks we face today. For example, we might be able to learn more about how to best coordinate to deal with risks collectively by seeing how people in small-scale societies proactively work together to decrease the likelihood and magnitude of risks and potential future hazards.
I would be very interested to hear from all of you about how need-based transfers and risk pooling fit within the broader frameworks of EA. In my reading so far, I have found some things that seem similar to or resonate with these ideas. For example, the idea of cash transfers is certainly adjacent to need-based transfers (though cash transfers tend to be about long-distance helping rather than building local networks). Also, the larger goals of creating common knowledge about the risk that we face seems to be a point of intersection, and indeed it is more explicit in the EA community than it is in need-based transfer networks. Are there other points of connection or interesting topics to explore at the intersection of these ideas and the general approaches of EA?
Thank you in advance for sharing your thoughts and ideas; I am looking forward to talking with you all here and learning from this community.
Hi, I appreciated your research into this, welcome to the forum and the EA community!
About 2 “There is no obligation to repay. Giving to those in need should not create an obligation for them to repay”
It seems like there is no obligation to repay shortly after, or for a specific incident. No ‘debt’ But there is an expectation that they will give in a similar instance of need by the giver, yes?
I’m struggling to think of a good case to practice/apply this system. Especially because I’m not sure it scales to larger societies, nor am I sure how to create norms around “need” and “obligation” where none were before. Or could it be written as a formal agreement? This is far outside my area-of-expertise so I really don’t have experience from which to draw examples of places this could be applied.
One flaw I can think of is that disaster risks tend to be localized. So if there is a bad year for crops, or a wildfire: chances are it hits both participant in the risk pool at the same time.
Can you explain why or which cases this is better than insurance for managing risk?
Hi EcologyInterventions, thanks for the welcome, and here are some thoughts/responses to your thoughtful questions:
Yes, there is an expectation that a need-based transfer partner will give if they are in similar circumstances (i.e., in need) in the future. But it is different from a debt/credit system, there is no keeping of accounts and paying back in need-based transfer systems. In need-based transfer systems people build relationships which essentially function as informal insurance systems. As in insurance systems, transfers only occur if certain conditions of need are met.
As for applying need-based transfers to large scale societies, there are some interesting existing examples. For instance, water sharing in the desert southwest: in Phoenix different regions can have water shortages and they transfer water among each other based on need. And also airline seat transfer systems: If an unexpected/uncontrollable situation happens like bad weather or equipment failure, airlines have agreements to sell each other seats well-below market rates to help reroute travelers.
There are also elements of need-based transfers that map onto how many people understand friendship. In fact, in my lab (in work led by my former graduate student, Jessica Ayers) we found that one of the main things that people expect from friends is that they will provide help in times of need.
Need-based transfers aren’t really an alternative to insurance; they are a type of informal and decentralized insurance that can potentially help to fill gaps that market-based insurance can’t fill or doesn’t fill as easily. One of the big advantages of need-based transfer systems is that they don’t require a lot of data (e.g., actuarial tables) in order to be viable for helping people manage a variety of risks. Instead, people build relationships with others based on an understanding of overall risk profiles, and generally agree to help each other in times of need.
You rightly point out that disasters can be localized and this can be a limitation of need-based systems. But interestingly, in our work in The Human Generosity Project, we see that people build these relationships often with people with different ecological conditions/risk. For example, the Maasai will have Osotua partners (Osotua is the Maasai need-based transfer system) who live in different regions. This helps to buffer against this risk of correlated shocks. Fijians, who deal with risks from cyclones (which often hit everyone in a village at the same time), build relationships among villages where whole villages essentially enter into these risk pooling/need-based transfer relationships with each other. We’ve written about these societies and their risk pooling systems in this chapter. We have also modeled correlated shocks in this paper if you’re interested in reading more about the technical side of these systems.
One other point of interest is that these need-based transfer relationships can start simply with a request for help. This is how it works among the Maasai, for example. If help is given after a prospective partner makes a request, then it starts a relationship of mutual obligation to help in times of need. It may very well work like this with some friendships in large-scale and market-based societies as well, with requests for need (and fulfillment of those requests) catalyzing new friendships that might function to help manage risk, even if people aren’t explicitly thinking about them in that way.
Thanks again for your great questions!