More on Intellectual Property Theory and Water Wells
My post on intellectual property theory, prompted a thoughtful email from David Ballantyne, who writes
Volokh hypothesises a water well which, while excludable, has such
capacity that a collection of farmers will make nonrivalrous use of the
well. In making this hypothesis, it is necessary to apprehend two types
of capacity: the total subsurface water available in the well, and
limits on the rate of extraction of water inherent in the well itself.
Precisely, the hypothesis is that the farmers have water needs which
will neither drain the well nor cause rivalry in extraction of water.
It must be noted that, being finite, the subsurface water is ultimately
limited; if one farmer desired, he could conceivably drain the water
table which feeds the well. It might take a fantastic project to
achieve this exhaustion, for example some ridiculous and ultimately
wasteful irrigation system, but it is theoretically possible.
Importantly, the exhaustion is subject to the desire of each of the
Similarly, the extraction capacity could be used up in a manner which
caused rivalry. One farmer could set up a pumping system which very
effectively transports water to his farm, but which restricts the others
from obtaining water. Or the pumping system could be entirely
ineffective; this is not at issue, the issue is that should a farmer
desire to make such a use of the well it would cause rivalry with the
other farmers, and that this is an inherent property of any physical well.
It is useful now to compare this with another hypothetical good: the
excludable, nonrivalrous TV remote control. Let's say that two
roommates have a single television with a single remote control. In our
hypothesis, neither of the roommates wish to watch television. In this
case, the remote control is nonrivalrous. Right? But clearly this is
untrue. Instead, the TV remote is rivalrous whether either roommate
wishes to watch TV at some particular time or not. The rivalrousness of
the TV remote comes about because _if the roommates both desired the
use_, rivalry would occur. It is no good hypothesising that this desire
does not exist - this does not affect the inherent rivalrous nature of
the TV remote.
This analysis is as relevant to the water well. Assuming that no farmer
desires to use so much water that the well is drained, or extract so
much water at a time or in such a manner that the extraction capacity of
the well is reached, does not change the fact that the total amount of
water is fundamentally limited, or that the extraction capacity of the
well is fundamentally limited. These things are necessarily rivalrous.
One might respond to this reasoning by proposing that the hypothesis be,
instead, that the water in the well is infinite, and that there are no
limits to the amounts or methods of extraction of the water. In this
case, the well is truly both excludable and nonrival. It is also now
unlike any physical object. In particular, given that we allow that we
can exclude citizens from trading information, this peculiar sort of
well acts exactly like, say, the collection of information which we call
an mp3. The argument now seems to be that we should consider the
hypothetical nonrivalrous well, which we might as well just call an mp3,
to be property solely because doing so would provide incentives for its
initial construction, and not due to any similarity with physical
property. That is, the question is solely one of pragmatic soundness,
with no direct analogy to inherently rivalrous physical property possible.
This conclusion comports with the common intuition among technophiles:
that analogies between physical property and intellectual property are
misleading, that any granted privilege to restrict trade and use of
information must be based on first principles without recourse to the
characteristics of physical property, and that using the word property
at all to describe such privileges is suspect.
posted by Lawrence Solum 3:12 PM