TCS Daily

From Here to Eternity?

By Arnold Kling - September 13, 2002 12:00 AM

In part one, I explained a vision for a new telecommunications architecture, consisting of "Packet Express" to deliver information signals and "Thingies" to convert those packets into sound, text, or video as desired. In this essay, I talk about the challenge of getting from here to there.

From There to Here

Compared with this relatively simple vision for the future, the current state is rather a mess. Most of our legacy devices are not as intelligent as Thingies. For example, a car radio does not understand packets. It only understands radio signals, and it only knows that something is a radio signal if it is broadcast within a particular frequency range. A traditional landline phone only recognizes phone signals. Moreover, it relies on the phone line to send it only phone signals, not other signals. Televisions are designed to pick up only television signals. And so on. Each type of device requires a separate network, in order to keep the signals straight.

Our legacy devices that require specialized networks were designed before the cumulative effects of Moore's Law were achieved. If we were starting from scratch today, given the progress that has taken place in computer chips, we would design all of our electronic devices as Thingies that interface with Packet Express, rather than as devices that respond only to specific types of signals on specific frequencies or lines.

If we had in place the simple architecture of Packet Express and Thingies, there is no way that new entrants could come in and introduce anything like our current TV's, phones, car radios, and so forth. These specialized devices, each requiring a different network for transmitting signals, would be way too expensive and inflexible.

Will Phone Companies Be an Obstacle?

Just because there is cheaper and better than here does not mean that it will be easy to get from here to there. In theory, incumbent telephone companies could try to thwart innovators that attempt to build the Packet Express infrastructure. The tendency will be for phone companies to make it difficult for new entrants to compete. In particular, phone companies will be inclined to use bundling as a weapon.

One economic model of bundling, described here among other places, is that it allows a monopolist to engage in price discrimination. As a monopolist, you would like to charge as high a price as possible, but offer discounts to your most price-sensitive customers in order to retain them. But how do you identify the customers who are price-sensitive, so that you only offer them discounts without offering discounts to those who are willing to pay higher prices?

One tactic for price discrimination is bundling. If a local phone company wants to maintain a high price for an individual service, such as plain old telephone service (POTS), while retaining price-sensitive customers, it can offer a bundle of services as an alternative. Consumers who obtain the bundle get discounts, while consumers who buy only one service pay full price. Some price discrimination is achieved, as consumers who very much want one service get singled out for higher charges.

A telecommunications upstart attempting to offer wireless Internet service could face destructive competition from the local phone company. The local phone company may be able to offer wireless at a below-cost price, as part of a bundle with its traditional service. Under this scenario, the local phone company makes up for the wireless subsidy embedded in its bundle by charging higher prices to its POTS customers. The upstart, not having a bundling option at its disposal, could be unable to match the local phone company's subsidy to wireless customers.

From society's point of view, this sort of bundling is difficult to evaluate. On the positive side, it would increase the consumer's incentive to switch to wireless Internet service, in order to take advantage of the bundling discount. However, this bundling might inhibit the growth of competition. It is difficult to say whether the net effect will be to accelerate or slow the deployment of Packet Express.

The Main Drag

The strategic behavior of local phone companies has the potential to slow the rate at which we change over to Packet Express. However, the main drag on the process is that so many of us have legacy devices, rather than Thingies.

Today, there are only a few million devices - newer laptop computers with wireless antennas - that are designed to work with Packet Express as it is envisioned here. Meanwhile, there are hundreds of millions of older computers, cell phones, car radios, and other devices that depend on the old architecture.

To get from here to there, American consumers will have to spend billions of dollars on Thingies. Eventually, they will get a return on this investment in the form of higher-quality services at lower cost. However, the manufacturers of Thingies will have to demonstrate the benefits before consumers will undertake the conversion.

Faced with consumer inertia, one might be tempted to look for a government solution. Unfortunately, government is more likely to be part of the problem. Just the other day, the FCC mandated that new televisions be able to receive signals in a new digital standard. That standard perpetuates the architecture of segregated communications networks, rather than Packet Express. It is as if at the dawn of the automobile era the government were issuing mandates for horse carriages.

Eventually, unless government policy serves to thwart the emerging architecture, consumers will increase their purchases of Thingies that are designed to work with Packet Express. As more Thingies populate the market, service providers will have an incentive to build the Packet Express network. Ultimately, the new architecture will win, because of these factors:

  • Lower cost of providing services. Your total costs for subscription television, telephone service, Internet access, cell phone service, and will be less than what you pay for those services today.

  • The ability to obtain any service, anywhere, at any time. Today, we only approach this with cell phone service. But we want ubiquity for email, music, and Web services as well.

Without any government mandates, Intel seems to be pushing in the direction of this new architecture. Their radio on a chip could be the core technology for making the new generation of electronic devices. Eventually, this will eliminate the main drag that is holding back conversion to the architecture of Packet Express and Thingies.



TCS Daily Archives