We should talk less about public goods funding and more about open source funding
2025 Mar 29
See all posts
We should talk less about public goods funding and more about open source funding
One topic that has been dear to me for a long time is the question of
how to fund public goods. If there is a project that provides
value to a million people (and there's no fine-grained way to choose who
gets the benefit and who doesn't), but each person only gets a small
amount of benefit, then it's quite possible that no single person will
find it in their interest to fund the project, even if the project is
extremely valuable overall. The language of "public goods" has a
century-long heritage in economics.
In digital ecosystems, especially decentralized digital
ecosystems, public goods are extremely important: in fact,
there's a strong case that the average good that someone might want
to produce is a public good. Open source software, academic
research into cryptographic and blockchain protocols, openly available
education resources, and many more things are all public goods.
However, the term "public good" has major challenges.
Particularly:
- The term "public good" often is used in public discourse to mean "a
good that is produced by a government", even if it is not a public good
in an economic sense. This leads to confusion, as it creates a
perception that whether or not a project is a public good is not a
function of what the project is and what its properties are, but rather
a function of who is building it and what their self-described
intentions are.
- There is a general perception that public goods funding lacks rigor
and is run on social
desirability bias - what sounds good, rather than what is good - and
favors insiders who can play the social game.
To me, these two problems are related: a big part of the
reason why the term "public good" is vulnerable to social gaming is
precisely the fact that the definition of "public good" is stretched so
easily.
Let's see what happens when you search
for the phrase "building a public good" on Twitter. I did this right
now, and here are some of the first results:
You can keep scrolling and find many projects using the phrase "we're
building a public good" to describe themselves.
The point of this is not to criticize the individual projects; I know
little about both of the above and they may well be excellent projects.
However, both of these examples are commercial projects that have their
own tokens. There is nothing wrong with being a commercial project, and
there is often nothing wrong with launching your own token. However,
it says something about the term "public good" when it so easily
gets diluted to the point where, today, it often seems to just mean
"project".
Open source
As an alternative to "public goods", let's think about the phrase
"open source". If you think about some central examples of things that
are clearly digital public goods, you will find that they are all open
source:
- Academic blockchain and cryptographic protocol research
- Documentation, tutorials...
- Open source software (eg. Ethereum clients, software
libraries...)
And on the flip side, open source projects seem to be by default
public goods. You can certainly come up with counterexamples: if I write
a piece of software that is heavily tailored toward my personal
workflow, and I put it up on github, the majority of the value created
by the project may still accrue to me personally. However, the act
of open-sourcing (as opposed to keeping it private) is certainly a
public good with very diffuse benefit.
A really nice thing about the term "open source" is that has
a clear and well-agreed definition. The FSF's Free
Software Definition and the OSI's Open Source Definition have stood
for decades, and there are natural ways to extend these definitions to
fields of endeavor other than software (eg. writing, research). In the
crypto space, the inherently stateful and multi-party nature of
applications, and the new vectors of centralized fragility and control
that these things imply, do mean that we need to extend the definition
somewhat: open standards, the insider attack
test and the walkaway test introduced
in this post can be a valuable addition to the FSF + OSI
definitions.
So what is the difference between "open source" and "public goods"?
Well, we can start off by asking the bot for examples:
I personally simply disagree with the claim that the examples in the
first category are not public goods. A project having a high barrier to
entry to contribute does not preclude it from being a public
good, and neither does corporations benefiting from the project. Also, a
project can absolutely be a public good while things around it are
private goods.
The second category is more interesting. First of all, we should note
that all five examples are in physical space, rather than
digital space. Hence, if we want to focus on digital
public goods, the above examples give no reason to oppose just focusing
on "open source". But what if we do want to also cover
physical goods? Even the crypto space has its share of enthusiasm for
better governing physical things and not just digital things; in a
sense, that's the whole point of network
states.
Open source and
physical local public goods
Here, we can make an observation: while providing these things at
local scale is an "infrastructure building" problem, and can be
done open source or closed source, the most efficient way to provide
these things at global scale generally ends up involving...
actual open source. Clean air is the most obvious example: there has
been lots
of research
and development,
much of it open source, to help people worldwide enjoy cleaner air. Open
source can help make any kind of public infrastructure easier to deploy
worldwide. The problem of how to provide physical infrastructure at a
local scale effectively is still important - but the problem applies
equally to democratically run communities and to corporations.
National defense is an interesting case. Here, I would argue the
following: if you build a project for a national defense reason that you
would not feel comfortable open-sourcing, then chances are that
while it may be a public good at the local scale, it's likely not a
public good at the global scale. Innovation in weapons is the most
obvious example. Sometimes, one side in a war has a much stronger moral
case than the other, and it's justified to help it with offensive
operations, but on average, building technology to improve military
capabilities does not improve the world. The exceptions
(national-defense projects that one would want to open source)
would likely be "defense" capabilities that are actually about
defense; one example might be decentralized agriculture,
electricity and internet infrastructure that can help people stay fed,
functional and connected in challenging environments.
Hence, here too it feels like shifting focus from "public goods" to
"open source" is actually the best thing to do. Open source should not
mean "it's equally virtuous to build whatever as long as it's open
source"; it should be about building and open-sourcing things that are
maximally valuable to humanity. But distinguishing which projects are
worth supporting and which projects are not is already well-understood
to be the primary task of public goods funding mechanisms.
We should talk less about public goods funding and more about open source funding
2025 Mar 29 See all postsOne topic that has been dear to me for a long time is the question of how to fund public goods. If there is a project that provides value to a million people (and there's no fine-grained way to choose who gets the benefit and who doesn't), but each person only gets a small amount of benefit, then it's quite possible that no single person will find it in their interest to fund the project, even if the project is extremely valuable overall. The language of "public goods" has a century-long heritage in economics. In digital ecosystems, especially decentralized digital ecosystems, public goods are extremely important: in fact, there's a strong case that the average good that someone might want to produce is a public good. Open source software, academic research into cryptographic and blockchain protocols, openly available education resources, and many more things are all public goods.
However, the term "public good" has major challenges. Particularly:
To me, these two problems are related: a big part of the reason why the term "public good" is vulnerable to social gaming is precisely the fact that the definition of "public good" is stretched so easily.
Let's see what happens when you search for the phrase "building a public good" on Twitter. I did this right now, and here are some of the first results:
You can keep scrolling and find many projects using the phrase "we're building a public good" to describe themselves.
The point of this is not to criticize the individual projects; I know little about both of the above and they may well be excellent projects. However, both of these examples are commercial projects that have their own tokens. There is nothing wrong with being a commercial project, and there is often nothing wrong with launching your own token. However, it says something about the term "public good" when it so easily gets diluted to the point where, today, it often seems to just mean "project".
Open source
As an alternative to "public goods", let's think about the phrase "open source". If you think about some central examples of things that are clearly digital public goods, you will find that they are all open source:
And on the flip side, open source projects seem to be by default public goods. You can certainly come up with counterexamples: if I write a piece of software that is heavily tailored toward my personal workflow, and I put it up on github, the majority of the value created by the project may still accrue to me personally. However, the act of open-sourcing (as opposed to keeping it private) is certainly a public good with very diffuse benefit.
A really nice thing about the term "open source" is that has a clear and well-agreed definition. The FSF's Free Software Definition and the OSI's Open Source Definition have stood for decades, and there are natural ways to extend these definitions to fields of endeavor other than software (eg. writing, research). In the crypto space, the inherently stateful and multi-party nature of applications, and the new vectors of centralized fragility and control that these things imply, do mean that we need to extend the definition somewhat: open standards, the insider attack test and the walkaway test introduced in this post can be a valuable addition to the FSF + OSI definitions.
So what is the difference between "open source" and "public goods"? Well, we can start off by asking the bot for examples:
I personally simply disagree with the claim that the examples in the first category are not public goods. A project having a high barrier to entry to contribute does not preclude it from being a public good, and neither does corporations benefiting from the project. Also, a project can absolutely be a public good while things around it are private goods.
The second category is more interesting. First of all, we should note that all five examples are in physical space, rather than digital space. Hence, if we want to focus on digital public goods, the above examples give no reason to oppose just focusing on "open source". But what if we do want to also cover physical goods? Even the crypto space has its share of enthusiasm for better governing physical things and not just digital things; in a sense, that's the whole point of network states.
Open source and physical local public goods
Here, we can make an observation: while providing these things at local scale is an "infrastructure building" problem, and can be done open source or closed source, the most efficient way to provide these things at global scale generally ends up involving... actual open source. Clean air is the most obvious example: there has been lots of research and development, much of it open source, to help people worldwide enjoy cleaner air. Open source can help make any kind of public infrastructure easier to deploy worldwide. The problem of how to provide physical infrastructure at a local scale effectively is still important - but the problem applies equally to democratically run communities and to corporations.
National defense is an interesting case. Here, I would argue the following: if you build a project for a national defense reason that you would not feel comfortable open-sourcing, then chances are that while it may be a public good at the local scale, it's likely not a public good at the global scale. Innovation in weapons is the most obvious example. Sometimes, one side in a war has a much stronger moral case than the other, and it's justified to help it with offensive operations, but on average, building technology to improve military capabilities does not improve the world. The exceptions (national-defense projects that one would want to open source) would likely be "defense" capabilities that are actually about defense; one example might be decentralized agriculture, electricity and internet infrastructure that can help people stay fed, functional and connected in challenging environments.
Hence, here too it feels like shifting focus from "public goods" to "open source" is actually the best thing to do. Open source should not mean "it's equally virtuous to build whatever as long as it's open source"; it should be about building and open-sourcing things that are maximally valuable to humanity. But distinguishing which projects are worth supporting and which projects are not is already well-understood to be the primary task of public goods funding mechanisms.