Innovation and the Evolution of Market Structure for Internet Access in the United States

This article looks to explain "how and why" the US commercial internet access market structure evolved during its first decade. The author, who is very fond of metadiscourse (probably five MS pages of telling the reader what he "will show" before actually getting around to it), explains that he will take 1993, 1998, and 2003 as points in time, and examine the evolution among them.

1993: A Nascent Internet Access Market

Privatization of the Internet occurred accidentally at first: due to some vagueness in the governing policies, and later the federal charter for the NSF (better description in previous chapter), it was opened up for broader access and a wider array of uses.

What with the genie out of the bottle, the NSF implemented a privatization plan to accommodate this growth, largely handing over many responsibilities to private corporations that were spun off from governmental and academic institutions.

Internet Service Providers would each maintain their own network, connecting their users to the backbone of the internet. Initially, each ISP network connected to a hub on the internet, but as congestion increased, national-level providers began to create their own networks by running parallel pipes to the public backbone wires to offer higher quality service to their customers (as a means of competitive advantage).

From 1993 to 1995, it was largely a small-scale phenomenon among academics who had used the internet for their work and computer enthusiasts who used it for their hobby. It largely went unnoticed by the corporate sector until 1994, when Netscape had its IPO, and curiosity was piqued.

In the early days, access to the residential user was via telephone, an industry that had been deregulated by the federal government, and many restrictions were still in place that prevented a local company from intruding on the territories of other local firms, and required each phone company to be nondiscriminatory in its pricing for local calls (charge all customers a flat fee, not pay-per-call). Fundamentally, the local telephone companies were pretty well tied up by this, and forced to play a very unprofitable role, but a very necessary one at this point in the Internet's evolution.

Access grew rapidly. By 1998, the major ISPs had reached all the major population centers in the US, with only 5% of the population being located in rural areas that had no access providers. However, 92% of the population had a choice of seven different ISPs.

It is also reasoned that many in the computing and communications industries failed to grasp the potential of the Internet. Very few commercial products were available, in terms of computers and software, and much of what was developed was hacked together and offered as freeware.

There were a few entrepreneurs who saw the coming storm, but they were independents rather than institutional, which led to the creation of hundreds of start-up companies offering niche products.

Simply stated, the big dogs of computing and communications totally missed the boat. Many of them struggled desperately - buying out smaller companies or launching internal efforts to get into the game, often competing with and taking an adversarial relation toward the institutions and community they were seeking to infiltrate - and as a result, they did not have much success, and were left far behind the pack.

Meanwhile, a number of companies were created by the academic sector - sometimes directly (created by a university) or accidentally (a group of students without faculty support). In many of the early battles, such groups won out over commercial competition: Netscape over Microsoft, Yahoo over various commercial search engines, etc.

1998: An Era of Impatience

By 1998, it was clear that the Internet was not merely a fad, but was a major new medium that would gain ubiquity in a very short time.

Broadband access began to emerge at this time to satisfy demands for content richer than plain text, requiring an ISP to provide a faster connection to the curb, and a more powerful network of their own. At this point, larger communications companies, which already had the infrastructure to provide high-speed access to commercial customers, regained share in the residential market.

Internet access began a trend which eventually leads to the disappearance of smaller ISPs. Some were beaten by larger firms, some were bought out, and others banded together to become a larger firm. Whatever the case, the era of smaller players came to a close, and very few were left by its end.

There's a bit of news about AOL here, which transitioned from a dial-up modem service to the leading national Internet Service Provider through as series of marketing moves (flooding the market with free trail offers) and savvy acquisitions. AOL catered to the neophyte - providing a "walled garden" of content inside the Internet jungle. This, ultimately, was AOL's downfall as well: users soon outgrew their need for ease-of-use, and found the walls of the garden to be more limited than protective (as they were unable to use tools and access information available to users of other ISPs), and having an "aol.com" e-mail account was a terrible embarrassment in an increasingly social medium.

On pricing, there had previously been various pricing models to the residential consumer - but competition in the market stabled pricing at a very specific point: about $20 per month for unlimited access via modem. This was largely a matter of price-matching, as this price represented the lowest that most companies could afford to charge without running in the red (and some providers even ran in the red, figuring that they could build a mass of customers and switch their cash flow by economies of scale). Interesting data point: AOL was the largest ISP, and one of the last to convert to a flat-rate pricing scheme - it was simply a price that the market insisted upon.

The number of backbone access providers also grew during this time - companies that developed high-speed connections between hubs, but did not sell Internet access to residential consumers. The commercial customers who were once served by the telephone companies and network providers drifted to this sector instead, where they could purchase greater capacity on a separate network than then residential customer.

In the regulatory sphere, there was little action at this time. Many of the groups that wished to govern the Internet (CERN, WWW Consortium) were private groups whose demands had no legislative teeth, or corporations without the ability to inflict its will on a market that had many options.

The Telecommunications Act of 1996 was a step forward to encourage open competition by forcing the larger telephone companies to share their networks with smaller providers. As a result, most companies divided their operations into separate subsidiaries: ILEC to manage their backbone network, and CLEC to provide a connection from that backbone to individual consumers, and used a bit of accounting wizardry to keep their networks inaccessible anyway.

During the late 1990's, there came a flurry of mergers, acquisitions, and reorganizations within the industry. While these were not attributed to legislation, federal agencies were often involved in the approval of such actions, and were generally very liberal in approving them.

The Internet Tax Freedom Act of 1998 was designed as a temporary measure to solve growing confusing over the application of federal and state taxes on Internet companies (both ISPs and retailers), and sought to ensure that aggressive local taxation would not inhibit the growth of a national medium.

2003: Reliability and New Transitions

By 2003, the Internet had come into its own as a medium, and access to the Internet had become ubiquitous in both the commercial and residential markets.

Broadband residential access grew rapidly, especially as local phone companies and cable television firms developed technologies to leverage existing connections to households (rather than having to run an entirely new data line) and competition decreased the cost of high-speed connections. Broadband access grew to 40% in urban and suburban areas by 2003 (though notably less in rural areas), and while prices for broadband were more flexible, they tended to be controlled by competition in local markets.

The author makes some mention of Wireless and mobile access. It was anticipated with much enthusiasm as early as 2001 that individuals could be provided access to mobile devices (cell phones, notebook computers), but it didn't catch fire at the time. Arguably, it is still in its infancy, and is struggling with the primitive capabilities of wireless devices.

The author mentions the growth in the value of computer stocks, rapid in the late 1990's but far more flat afterward. He doesn't, however, mention the dot-com crash that began in 1999 and ran through 2001. His explanation for the flattening is the decrease in the year-to-year expansion of the market as adoption nears the 100% mark - which is more in line with traditional views of economic growth.

He speaks a bit more on the consolidation of access providers, but this is merely the continuation of the trend from the previous section.

He also mentions this as a period of consolidation online, from a broad array of small vendors to a few large ones in most product categories, as a result of competition.

Summary of the First Decade

While the rise of the Internet and its evolution into a mature industry within its first decade seems dramatic, it largely follows patterns seen in most other industries. There are a few quirks - in that the Internet often evolved by accident or coincidence rather than by a managed and deliberate process - but in most cases, they have been temporary in nature: the market adjusted, or is in the process of adjusting, to a normal course of industrial evolution.