Written by Ivan Vendrov and Mary Stowers. Disclaimer: neither of us has formal economics training so don’t rely on our descriptions of economic concepts being very accurate.
Introduction
From the Coasean perspective, almost everything bad in the world is due to high transaction costs. This comes from Ronald Coase’s famous theorem, which states roughly that if transaction costs are sufficiently low, we get a Pareto-efficient outcome (a world where all possible win-win bargains have been exhausted, such that any change results in at least one person being worse off.)
Taking the contrapositive, any observed Pareto inefficiency in the world is due to high transaction costs. But what are transaction costs exactly? There are a number of differing definitions in economics literature, but the broadest one is “anything that gets in the way of signing contracts between willing participants”. (This broad definition makes the Coase theorem somewhat trivial.) Generators of transaction costs include:
Inability to perfectly commit to a mutually beneficial contract, due to a lack of mutual trust.
Difficulty of measuring performance / adherence to contract
Negotiation costs from specifying what should happen in every possible contingency.
Communication and negotiation costs of creating contracts with large numbers of participants, such as contracts affecting the environment, common resources, or standardization within an industry.
Poorly defined or insecure property rights making voluntary contracts unenforceable or undesirable relative to a violent or political course of action.
Asymmetric information leading to missing markets because lower-information participants fear being exploited, e.g. in the used car market as described in the classic paper The Market for Lemons.
The Transaction Costs approach to EA
If we take this model seriously, it suggests that instead of helping people directly, it’s much more impactful to reduce transaction costs so they can help themselves and each other by negotiating positive-sum bargains. The image we have in mind is: people’s energy and drive to improve their lot in life is incredibly strong, and transaction costs are a master variable in how that energy gets directed. When transaction costs are high, the energy is dissipated in wasteful zero-sum competition. When transaction costs are low, that energy is channeled into myriad positive-sum interactions.
For example, it’s been estimated that a dollar spent on education in very poor countries increases a child’s lifetime income by $10 or more. You could try to address this problem directly by funding education, but the Coasean view suggests that the real problem is a missing market due to high transaction costs. Why isn’t anyone capitalizing on an opportunity to 10x their money? An investor could sign an income sharing agreement where they fund a student’s education in exchange for some modest fraction of their income over the next ten years. Notably, this investor would now have a strong incentive to improve this student’s life in all sorts of ways, such as nutrition, health, and helping connect to job opportunities. However, for some reason, this contract cannot be signed; for one thing, it may be illegal- often for good reason, because signing long-term contracts with minors is ethically fraught and similar indentured servitude contracts have been used in an exploitative way in the past.
On the other hand, a dollar spent on improving people’s ability to strike such long-term contracts may result in much more impact than a dollar spent on helping people directly, due to the huge number of positive-sum bargains it unlocks.
Morally Relevant Non-Participants
The biggest issue with the transaction cost approach to EA is that many morally relevant actors can’t participate in current markets or deals at all, such as non-human animals and far future people. Is there a way to reduce transaction costs with them? Focusing on far future people in particular, it doesn’t seem like there are any win-win bargains to be struck because there’s nothing that people who don’t yet exist can offer us, so the whole Coasean analysis fails.
However, there are some organizations, such as states, universities, or nonprofit foundations, that explicitly have long (100+ year) time horizons. Many people also care, at least somewhat, about what happens to the world in the long term future. Can we formalize their claims on the future in some way, and make it easier for them to trade with each other and with more short-termist actors?
As a silly example, imagine we sold all the polar bears. Specifically, if there are any polar bears alive in 100 years, they all become the property of the Polar Bear Foundation. The Polar Bear Foundation would now be strongly incentivized to preserve polar bears and their ecosystems and habitats in a way that no individual country or organization currently is. More generally, we can do this for any important species, ecosystems, or non-renewable resources that are currently threatened or endangered.
Other directions to reduce transaction costs with longtermist actors include:
Selling financial instruments that only pay out in the very long term, incentivizing longtermist individuals and foundations to buy them, take actions that help the long-term future, and sell them at a profit.
Increasing the stability and security of property rights. How could humanity credibly commit to respect a property right for 1000 years, no matter what environmental changes, wars, and technological changes come in between?
Institutions like the Long Term Stock Exchange, that make it easier to create organizations optimizing for the long term.
What about improving the lot of nonhuman animals? Even people heavily invested in factory farming understand that it’s likely to be judged extremely unethical by people in the far future. The Coasean perspective suggests this means there’s a bargain to be struck. Indeed, we could facilitate a bargain with these far future people by issuing certificates of impact to anyone who reduces factory farming, whether by direct action or by improving clean meat technology. We expect far future people to value these certificates highly because they’ll be much richer and more concerned about non-human animals than we are; this increases the value of these certificates even in the short term, and encourages people to work harder on eliminating factory farming even if they are purely interested in profit.
Summary
The Coase theorem suggests that reducing transaction costs is a promising and underrated way to improve the world across three of EA’s core focus areas: improving the lives of the worst off people now, helping reduce the suffering of non-human animals, and making the long-term future go well. We described a number of directions and ideas, but the problem of figuring out what the key transaction costs are, and finding the most effective ways to reduce them, remains almost completely open.
Reducing Transaction Costs, or Coasean Effective Altruism
Written by Ivan Vendrov and Mary Stowers. Disclaimer: neither of us has formal economics training so don’t rely on our descriptions of economic concepts being very accurate.
Introduction
From the Coasean perspective, almost everything bad in the world is due to high transaction costs. This comes from Ronald Coase’s famous theorem, which states roughly that if transaction costs are sufficiently low, we get a Pareto-efficient outcome (a world where all possible win-win bargains have been exhausted, such that any change results in at least one person being worse off.)
Taking the contrapositive, any observed Pareto inefficiency in the world is due to high transaction costs. But what are transaction costs exactly? There are a number of differing definitions in economics literature, but the broadest one is “anything that gets in the way of signing contracts between willing participants”. (This broad definition makes the Coase theorem somewhat trivial.) Generators of transaction costs include:
Inability to perfectly commit to a mutually beneficial contract, due to a lack of mutual trust.
Inability to fully specify such a contract (incomplete contracting)
Difficulty of measuring performance / adherence to contract
Negotiation costs from specifying what should happen in every possible contingency.
Communication and negotiation costs of creating contracts with large numbers of participants, such as contracts affecting the environment, common resources, or standardization within an industry.
Poorly defined or insecure property rights making voluntary contracts unenforceable or undesirable relative to a violent or political course of action.
Asymmetric information leading to missing markets because lower-information participants fear being exploited, e.g. in the used car market as described in the classic paper The Market for Lemons.
The Transaction Costs approach to EA
If we take this model seriously, it suggests that instead of helping people directly, it’s much more impactful to reduce transaction costs so they can help themselves and each other by negotiating positive-sum bargains. The image we have in mind is: people’s energy and drive to improve their lot in life is incredibly strong, and transaction costs are a master variable in how that energy gets directed. When transaction costs are high, the energy is dissipated in wasteful zero-sum competition. When transaction costs are low, that energy is channeled into myriad positive-sum interactions.
For example, it’s been estimated that a dollar spent on education in very poor countries increases a child’s lifetime income by $10 or more. You could try to address this problem directly by funding education, but the Coasean view suggests that the real problem is a missing market due to high transaction costs. Why isn’t anyone capitalizing on an opportunity to 10x their money? An investor could sign an income sharing agreement where they fund a student’s education in exchange for some modest fraction of their income over the next ten years. Notably, this investor would now have a strong incentive to improve this student’s life in all sorts of ways, such as nutrition, health, and helping connect to job opportunities. However, for some reason, this contract cannot be signed; for one thing, it may be illegal- often for good reason, because signing long-term contracts with minors is ethically fraught and similar indentured servitude contracts have been used in an exploitative way in the past.
On the other hand, a dollar spent on improving people’s ability to strike such long-term contracts may result in much more impact than a dollar spent on helping people directly, due to the huge number of positive-sum bargains it unlocks.
Morally Relevant Non-Participants
The biggest issue with the transaction cost approach to EA is that many morally relevant actors can’t participate in current markets or deals at all, such as non-human animals and far future people. Is there a way to reduce transaction costs with them? Focusing on far future people in particular, it doesn’t seem like there are any win-win bargains to be struck because there’s nothing that people who don’t yet exist can offer us, so the whole Coasean analysis fails.
However, there are some organizations, such as states, universities, or nonprofit foundations, that explicitly have long (100+ year) time horizons. Many people also care, at least somewhat, about what happens to the world in the long term future. Can we formalize their claims on the future in some way, and make it easier for them to trade with each other and with more short-termist actors?
As a silly example, imagine we sold all the polar bears. Specifically, if there are any polar bears alive in 100 years, they all become the property of the Polar Bear Foundation. The Polar Bear Foundation would now be strongly incentivized to preserve polar bears and their ecosystems and habitats in a way that no individual country or organization currently is. More generally, we can do this for any important species, ecosystems, or non-renewable resources that are currently threatened or endangered.
Other directions to reduce transaction costs with longtermist actors include:
Selling financial instruments that only pay out in the very long term, incentivizing longtermist individuals and foundations to buy them, take actions that help the long-term future, and sell them at a profit.
Increasing the stability and security of property rights. How could humanity credibly commit to respect a property right for 1000 years, no matter what environmental changes, wars, and technological changes come in between?
Institutions like the Long Term Stock Exchange, that make it easier to create organizations optimizing for the long term.
What about improving the lot of nonhuman animals? Even people heavily invested in factory farming understand that it’s likely to be judged extremely unethical by people in the far future. The Coasean perspective suggests this means there’s a bargain to be struck. Indeed, we could facilitate a bargain with these far future people by issuing certificates of impact to anyone who reduces factory farming, whether by direct action or by improving clean meat technology. We expect far future people to value these certificates highly because they’ll be much richer and more concerned about non-human animals than we are; this increases the value of these certificates even in the short term, and encourages people to work harder on eliminating factory farming even if they are purely interested in profit.
Summary
The Coase theorem suggests that reducing transaction costs is a promising and underrated way to improve the world across three of EA’s core focus areas: improving the lives of the worst off people now, helping reduce the suffering of non-human animals, and making the long-term future go well. We described a number of directions and ideas, but the problem of figuring out what the key transaction costs are, and finding the most effective ways to reduce them, remains almost completely open.