The Economic Efficiency of Internet Public Goods

This chapter is an article written in advance of the NSF's opening the internet to commercial interests, which argued that the free market would be "an appropriate mechanism to drive the future development of the internet" in the fact of those who assumed that the commercialization of the internet would be harmful to its potential to be an "equalizer" in the information economy.

Economic Definitions: Public Goods, Private Goods

By definition, a public good is one that is not depleted by use: the fact that one person uses it does not make it unavailable to others, and benefits those that do not pay for it directly. The example given is national defense: this protects all citizens regardless of the amount they have paid in taxes.

By contrast, a private good is depletable and excludable: when one person eats an apple, another person cannot eat the same apple, and the world supply of apples is depleted by one.

An "externality" is defined as a public good that arises as a side effect of the creation of a private good. An example is education: a college provides an education to its students (private good), but the externality is the proportion of these students who remain in the community and improve its overall economy.

The assumption that the Internet is a public good is a fallacy: while it is presently financed by the federal government, bandwidth is both depletable and excludable, in that the network capacity used by one individual depletes its availability to others. And since access to the network is limited to specific physical locations, it cannot be argued to be of benefit to anyone who is not granted access.

Software as Public Goods

The author turns to a specific piece of software that was developed at a university, then placed on its server for others to download at no cost.

Said another way, it was freeware: the only obligation under the software license was that anyone who modified the software was required (or "requested," as such clauses had no legal teeth in those days) to provide their modifications back to the producer, so that it could be made available to the community.

The author suggests that this is the model of a public good: the development cost was underwritten by the university, as the software addressed a specific need. However, once developed, it could be downloaded and installed by others at no cost to the consumer - and it could be provided at no cost to the supplier.

The drawback to providing the program to the public in this manner is that the university raised no funds from the product, so it was unable to generate any funding to increase the development staff resources in order to deliver critical enhancements to the product - and as a consequence, the software did not evolve to meet the changing needs of its customer base - hence the product was eventually replaced by a commercial version.

Computer Networks as Public Goods

Access to a computer network seems like a public good: once the cost to set up the network is spent, the cost of adding an additional user is zero. However, this remains true until the network has reached its capacity, and congestion occurs, rendering the network less usable, and ultimately unusable.

At that point, an additional infusion of funds is necessary in order to evolve the network to deliver additional capacity for all users - but so long as access is free, there is no funding to cover the costs of expansion. Like the freeware public good described above, it will eventually be replaced by a commercial version.

The University Model and the Worldwide Knowledge Industry

The University is, in effect, a knowledge mill: scholars are required to "publish or perish," creating an ever-growing body of knowledge, and that knowledge is largely considered to be a public good.

However, the university system is itself a depletable and exclusive resource: to benefit from the knowledge created at universities, one must have access to it; and to have access, one must pay tuition. As a result, education is a private good.

Even if it were not so, and anyone could wander onto a university campus to sit in on any lecture, there are a finite supply of seats in the university auditorium. While it would not be a private good, neither would it be a public one - though some might argue that it is an "impeded" public good (only the limitation of seats prevents an infinite number of people from attending a lecture).

The Internet is hailed by many as the solution to this problem: by making knowledge available electronically, in a format such that one person's consumption of it does not diminish the source, nor prevent the use of the same knowledge by others.

However, given the nature of the computer network, and the congestion that will inevitably occur, this is not so: while the knowledge is undiminished by downloading it, the bandwidth to download it is not.


Given that internet capacity is a depletable resource, we are faced with four options:

  1. Do nothing - Treated as a public good, the network will continue to operate until it reaches its capacity, and eventually becomes utterly unusable
  2. Reinvest University Resources - Continue to divert funds from the university's coffers to extend network capacity, reducing the budget to create new knowledge for the sake of providing access to old
  3. Taxation - Turn to government sources to obtain additional revenues in taxes to support the network for those individuals who have access
  4. Privatization - Transition the Internet into a private good, requiring users to pay for access, depending on those who sell access to reinvest a portion of their revenue into increasing network capacity

The author is inclined to support the fourth option: network capacity is a private good, and should be treated as such, relieving universities of the burden of supporting the networks instead of creating knowledge, while ensuring that the Internet can serve as a medium through which knowledge itself can be unimpeded as a public good.