Internet Pricing: A Regulatory Imperative

Based on the rise of "bandwidth-hungry" applications, this author suggests that it is imperative to develop a usage-sensitive pricing system that will effectively "ration" bandwidth, as well as putting into regulatory legislation and agencies to ensure consumer welfare.

EN: This gets back to the fallacy of limited resources, plus I'm catching the aroma of a political agenda, but I'll continue annotating the article in an attempt to understand such mindsets.

The Different Dimensions of Growth

He briefly explores the growth of Internet traffic as multidimensional:

In all, this will lead to an exponential increase in demand that will overwhelm the network's capacity to transmit data, and he questions whether these new users, new purposes, and new media serve a greater good, or if they may overwhelm the network with data that serves less noble ends.

Pricing Alternatives

The flat-rate pricing offered to most customers encourages overuse and overconsumption. From a purely economic perspective, the customer who fails to utilize the full amount of bandwidth allotted to him, all the time, is gaining less utility for the same price.

A supplier-driven model (making companies that host information pay the costs for files people retrieve from its server) may be a solution, but it would have a negative impact on the medium: it would cause free information resources to dry up.

In any economic exchange, value is granted to receive value - and as information providers derive no benefit, they have no economic incentive to pay the cost. Instead, the benefit is enjoyed by the recipient, who should bear the cost of receiving the information.

Arguments Against Flat-Rate Pricing

The flat-rate model is based on covering fixed costs, as the marginal cost of providing access to an established network for an additional user is zero, so long as the bandwidth is sufficient to accommodate the additional traffic.

Even when the addition of users exceeds network capabilities, the supplier can add network capacity and continue to offer a flat-rate price, basing his estimation on a cost-per-user in bulk.

As bandwidth consumption increases, the network capacity 9number of cables and routers) needed per user will increase, and so must the price charged to consumers.

The economic injustice of this is in the difference in usage among consumers: a consumer who users little bandwidth is charged the same price as the one who consumes an excessive amount.

An analogy is drawn to electric power: the network and generating capacity are shared by all users, but unless cost is metered on the basis of usage, the individual consumer pays no tariff for his individual benefits, and is not inclined to use scarce resources in a conservative manner.

Hence, the responsible behavior of some consumers is penalized to subsidize the wasteful consumption of others.

The Costs of Congestion

The nature of the Internet is such that a single users can bog down a significant portion of the network with little cost or effort: the example is a university student who uses a $2,000 computer and a $500 camcorder to send a video stream that would hog the entire T1 connection to his campus, precluding other users from accessing the network for research purposes.

The problem is largely one of network design: since few users use the full capacity of their lines most of the time, the network is designed such that the convergence of user connections at the node is far greater than the capacity of the line from the node to the next hub on the network.

Pricing Models

There are various pricing models to apportion the cost of network capacity to the users who are consuming it.

The telephone model, which determines pricing by determining the distance of a connection and the route it travels from sender to receiver, is not feasible because packets are routed by capacity, not distance, so the packets between user and host do not take the shortest possible route, but the route that has the greatest capacity for the traffic at any given instant, which may be different routes at different times.

The precedence model decides the level of priority based on media type, and could give low-bandwidth traffic (text) greater preference than high-bandwidth traffic (video), thus doing the most good for the greatest number of users given the limited resources. The limitation is that, without evaluating the content, mime type alone does not necessarily indicate the greatest social good (a thousand e-mails full of gossip would be considered more important than a video illustrating a medical technique).

A "smart market" mechanism would give the receiver of information the ability to indicate the price he was willing to pay to receive it promptly, and the highest "bid" (reflected in the headers) would receive greatest priority. While economically sound, this caters to commercial priority rather than scientific priority, such that the commercial value, rather than societal value of information would take precedence.

Building a Case for Regulation

Since none of the pricing models alone can provide a resolution to the bandwidth competition that is satisfactory (to the author), regulation by a disinterested third party is necessary to determine a manner of apportioning scarce resources to those with the greatest need, for the greatest possible social good.