Common goods vs public goods

From an email sent to me by Aaron Wolf of , posted with his permission:

Some quick thoughts on FNF and common vs public goods:

The non-rivalrous aspect of bandwidth isn’t a straightforward yes or no.
Just like a road where some vehicles (bikes, pedestrians) have
negligible impact and where sometimes there’s no issue of traffic
congestion, there are circumstances that are effectively non-rivalrous.

Perhaps there’s a case where the overall bandwidth of a network is such
that the first X users can all use as much bandwidth as they possibly
want without any rivalry. So, the resource is effectively a public good
as long as there’s less than X users.

In practice, for fundraising it works like this:

If there’s a threshold of funding that is required just for the system
to exist, then you need the users to pay at least that overall, so
nobody should be charged until that threshold is met. If there’s then
benefits to the system with additional funding, and everyone benefits
from that and the number of users is still less than X (from above),
then Crowdmatching (our new term for our funding model) makes sense. As
one user in that case, I want to say “hey, we all want the benefits of
more funding, and we all fully share the benefits, so I pledge to match
everyone else who chips in!”

But if we hit a level where I don’t really care about the benefits that
extra funding would bring or no further funding is needed, then we’re at
the public-goods-success level (i.e. the “we did it!” level) where we
can shift to focusing on sharing the burden fairly (i.e. if we’re still
less than X users, we could reduce each users’ donations when new
people are added to help out). I suspect it’s unlikely to get to this
place while still <X users.

Once we hit X users, then the public good becomes a common good. Now, I
am no longer happy to say “I will match every new donor” because even
though I appreciate some benefits of extra funding, my bandwidth is
hurt by the presence of more users, so I am not fully happy with
encouraging more users, at least not to the point that I want to match
them all. This is where managing the limited resource (bandwidth)
becomes trickier to solve if we keep things non-excludable.

You would know whether X users in this example is even a thing or if all
users have rivalry for bandwidth even from small number of users (or if
X is too small to be worth thinking about).

Finally, if increased funding also increases bandwidth for everyone
and is not just negligible for existing users but also has some
significant benefits, then X increases with more funding and so the
public-goods state covers a higher number of users in terms of the
funding model.

In short: Crowdmatching makes sense for public goods, not as much for
common-pool resources, but thinks that have scarcity in principle may be
non-rivalrous in practice, and it’s the practical rivalry or
non-rivalry that matters in terms of crowdmatching or not. Of course,
any excludability (either private or club goods) is squarely outside of
where crowdmatching would apply.

Hope that’s clear and helpful.