But a strong case can be made that no medium prior to the Internet has grown so quickly to touch the daily lives of so many people.

According to a 1999 study by The Strategis Group cited in CyberAtlas, 37 million Americans were using the Internet from home on a daily basis at the end of 1998 compared to only 19 million in the middle of 1997. Strategis also estimated that 34 million American households were using the Internet for a total of 65 million hours per day. A report on the Strategis study in CyberAtlas also indicated a dramatic surge in daily Internet use at the office from 19 million Americans in mid-1997 to 32 million at the end of 1998.

According to Inteco, 1998 was a year of critical breakthroughs for the Internet and the surge of users on the World Wide Web cannot entirely be attributed only to the increased popularity of online shopping during the last holiday shopping season.

The Internet revolution in the commercial, social, and civic life of America also appears to be accelerating technological innovation and the convergence of various technologies into an entirely new communications environment.

When governments do attempt to formulate policies for heretofore geographic-based commercial and social activity that is migrating to the Internet, they must do so very carefully and remain mindful of the tremendous complexities of a global interactive medium.

The U.S. Supreme Court has already set aside one congressional attempt to control content on the Internet because of the law’s impermissible restriction of the First Amendment guarantee of free speech. Governments overseas however, not bound by the U.S. Constitution or its First Amendment, do continue to try to find ways to control content.

This survey is based on existing research. The staff of the United States Internet Council has attempted to summarize key trends that emerge from the research of companies and academic institutions in one brief report.

We believe the trends we have observed indicate the following:

The studies we have seen do not suggest that rapid growth of this new medium might mean domination by any single technology. In fact, quite the opposite seems to be the case. High demand means that all technologies, old and new, are making a contribution to the rapid evolution of the Internet. For example, despite the emergence of ADSL, cable, satellite, terrestrial wireless, fiber optic and other access technologies that may yet come along, more than half of all Internet users will likely still gain access over copper wires five years from now in 2004.

Based on history, we believe that any decision governments might make now about which technology should be advanced will certainly be fraught with unintended consequences. The growth of the medium will not be helped in the long run by governmental policies that impose artificial distortions in the marketplace.

Internet Use Continues to Surge

Figure 1

Source: Network Wizards (1999)

A host is simply any computer that is directly reachable through the Internet. To do this, it must have a unique address and have the capacity of providing information in addition to access. For purposes of measuring the size of the Internet, a host is not a computer that is within an intranet and hidden by a firewall.

The 43.2 million hosts represent a 69 percent increase over the 29.6 million hosts in January 1998.

By every measure, Internet use continues to outstrip even optimistic predictions. In early 1993, about 90,000 Americans had access to the Internet. That number had grown to about 81 million in early 1999, an increase of about 900 percent. In September of 1997, Cyber Stats reported 43.6 million U.S. adults using the Internet during a 30-day period. By August of 1998, that figure had increased to 53.5 million U.S. adults.

Computer Industry Almanac found 61 million Internet users worldwide at the end of 1996, 148 million at the end of 1998, and the editors project a figure of 320 million for the end of the year 2000. While figures differ widely among various research companies using different methodologies, the trend is clear: Internet use is growing exponentially.

The growth in the number and percentage of households in the United States with Internet access provides another measure of Internet growth.

The chart in Figure 4 shows that the Internet grew from 5.8 million households in 1994 to the present 38.8 million, as reported by Forrester Research. Forrester goes on to project household access at almost 60 million households by the year 2003. These numbers reflect a growth in the percentage of all U.S. households that have access to the Internet from about 5% in 1994 to 33% today.

Forrester also projects that more than 50% of all U.S. households will have access by 2003.

E-mail is the most popular Internet activity for most people. According to a report in eMarketer, 84% of Internet users send and receive e-mail on a regular basis.

Here again, research estimates vary widely on e-mail usage depending on methodology. The Yankee Group reports 263 million e-mail mailboxes worldwide, but there are different estimates for the number of e-mails actually sent. Estimates range from a low of 618 billion messages in 1998 to as many as four trillion. That last number comes from an Iconocast report which forecasts a volume of more than seven trillion e-mails in 2000. And a 1998 Forrester Research report projects that 50% of the U.S. population, 135 million people, will use e-mail by 2001.

The traditional picture of Internet users as white males is no longer accurate. With each passing month, the population of Internet more closely resembles the demographics of the American people as a whole.

One of the persistent concerns about the Internet revolution is that minorities will be left behind. Those concerns have generated considerable political interest, including proposals for government regulation and subsidies. Actual evidence of minority participation in the online world is much more optimistic.

It is true that a poll conducted in 1997 by Lou Harris & Associates and Baruch College estimated the percentage of online users that were black or Hispanic was only 6% for each group. But Forrester Research recently found that Internet access for black Americans is today at 23%, and that it will reach 40% by the year 2000. Access for Hispanics is today at 36%, and it will reach 43% by 2000. Americans of Asian heritage are most likely to be Internet users. The Forrester study found 64% of Asian Americans are on the Internet today, and projects that by 2000, 68% will be. (See Figure 9 next page)

Research indicates that economic factors, such as the ability to afford a computer, the level of education, and attitudes towards technology are more important for determining who is using the Internet than ethnic status. These factors are, of course, correlated with racial and ethnic status, but as the economic disparities continue to narrow, Internet access will follow. There do not seem to be important cultural barriers to Internet use.

The same study found that in early 1996 about 60 percent of Internet users earned an income of $50,000 per year or more. This percentage decreased to 54 percent by the end of 1998. (See Figure 11 next page)

Forrester Research estimated in 1997 that residents of five million U.S. households had shopped for some product using the Internet. The number for 1998 was 10 million and the forecast for 1999 is that 13 million U.S. households will shop on the Internet.

At least one 1999 study indicates that e-mail is becoming the dominant mode of business communication and is becoming even more important as a business tool than the traditional voice telephone.

However, other modes of connectivity, such as advanced telephony systems, cable, wireless, and satellite, are beginning to emerge.

According to Jupiter Communications, there were about 15.1 million U.S. households connected to the Internet via copper wire in 1996. By 1998, 27.6 million were connected over copper and 500,000 were connected by advanced telephony systems such as ISDN and DSL, and another 500,000 homes were connected with cable.

By 2000, Jupiter predicts that 35.2 million households will connect with copper, 2.2 million will connect with advanced telephony, 2.9 million will connect via cable, and 500,000 will connect using wireless and satellite systems.

According to Professor Shane Greenstein of Northwestern University’s Kellogg School of Management, more than 67.5 percent of people in the United States live in counties that are served by more than 10 Internet Service Providers (ISPs) providing local dial up service. For these urban area residents, access to the Internet is very easy. The cost a customer incurs for the telephone call connection is usually the cost of a local call and usually does not include a long distance toll but may vary depending on the distance from the home to the ISP.

Seventy-eight percent of Americans live in counties with four or more choices of Internet service.

As Professor Greenstein points out, the analysis by county of residence may overstate the lack of connectivity.

"In many cases, residents of a county with no ISP may still be able to easily access an ISP in an adjacent area. The fact that there is no ISP located in a county does not necessarily mean that no low-cost service is available," Greenstein told the U.S. Internet Council on April 8, 1999.1 

The increasing popularity of the Internet as an interactive medium for commerce and entertainment is driving more demand for applications that require the availability of broader bandwidth. The number one factor that attracts visitors to return to web sites is the quality and availability of content. More and more of popular content requires graphics, audio, and video data streams. Such data intensive streams require large pipes to accommodate downloads of files. (See Figure 17 next page)

Internet service over a cable TV system is not currently a common offering for business locations. However, many businesses do access the Internet from office buildings connected to ISDN or ADSL lines. Prudential estimates that 94 million subscribers will be qualified for ADSL service by 2004.

In addition, there is just beginning to be an offering of fiber optic service to some office buildings in cities where a fiber optic loop is in place.

Many companies are building fiber optic loops around major cities and are offering high-speed and broadband Internet access to business customers.

For example, Hyperion, a company owned by Adelphia Cable in Pennsylvania, was offering high speed and broadband Internet connections to businesses in 46 cities on 22 fiber networks in early 1999. Other major players in constructing fiber loops and offering broadband Internet service include Time-Warner Telecom, RCN, Cox Fibernet, FMS, and Teleport, among others.

Other major companies are making huge investments in terrestrial wireless Internet service.

There are 49 content networks such as Talk America, Disney, and others that offer streaming audio on the Internet. There are 168 radio stations and at least one television station that broadcast only on the Internet and not over the airwaves. In short, a fledgling Internet-only broadcast medium is beginning to emerge.

Many Internet users say they want to be able to download broadband applications such as music videos and even longer feature films. In a few years, the Internet could become a video jukebox in the home and a home movie theater.

According to Real Networks, the first RealPlayer was released in 1995. By the end of 1998, there were over 50 million registered users. The download rate exceeds 175,000 per day, an increase more than 270 percent since the start of 1997. Every week, over 145,000 hours of live sports, music, news, and entertainment are broadcast over the Internet using RealSystem technology.

More and more people are playing computer games online, many of which have sophisticated graphics and complex interactive applications that require more bandwidth.

Federal, state, and local government threats to the Internet have taken a variety of forms and are often aimed at restricting or controlling content that many policy makers may find objectionable. Legislation by anecdote should be avoided. Policy planning based on a solid understanding of the non-geographic architecture and operating characteristics of the Internet can best be assured through active dialogue between industry leaders and policy leaders. Consumers are already voting everyday with the choices they make in the marketplace.

Congress passed the first version of the Communications Decency Act in 1996 as an attempt to protect children from obscene content. While protecting children from harmful content is an extremely worthy goal, the U.S. Supreme Court found that the measure impermissibly intruded on First Amendment guarantees of free speech. The role of government, as distinct from the role of parents and guardians, must be somewhat limited in overseeing the content that children see on the Internet.

Other Congressional attempts to impose ceilings on the export of encryption products and to adopt other versions of content control continue to be a detriment to the adoption of a stable legal platform for the expansion of electronic commerce.

On the plus side, Congress did in 1998 adopt a number of measures helpful to the high-tech industry that support the infrastructure of the Internet including measures to allow work permits for high-tech workers from abroad. In addition, a law passed by Congress and signed by President Clinton has created a study commission to wrestle with the problem of electronic commerce sales taxes and the states.

At the level of state governments, there is good news and bad news. The bad news is that some state legislatures in 1997 did debate bills to control content on the Internet. Some content control proposals have been limited to access from public schools but others have been much broader in scope in that they sought to affect the entire population of a state.

On the positive side, several states have moved in the direction of Virginia by debating bills to create state level Internet study commissions and policy plans. On the whole, state legislatures meeting in 1999 sessions have been far less apt to debate bills seeking to control content on the Internet that they were in 1997. As lawmakers gain a better understanding of the basic architecture of the Internet, better and less intrusive Internet policy has taken shape.

So far, at least, the Internet is living up to its billing. The Internet, itself, continues to grow very rapidly, as does the number of households online and the sheer volume of Internet content. At work and at home, Americans are using the Internet with increasing frequency to stay in touch and to do business. All data suggests that this growth will continue, and in many cases accelerate, well into the next century.

Already, the Internet has diffused throughout the United States faster than any previous communications technology. Inexpensive dial-up access to the Internet is available to the vast majority of Americans. Americans are beginning to demand increasingly robust connection to the Internet and this demand is beginning to be met by an array of new broadband technologies which span the converging telephone, cable, satellite and wireless industries.

As the Internet becomes more central to the way we live, work and play, it becomes a topic of increasing interest to policymakers at all levels of government. The data assembled here suggests three imperatives for government officials as they grapple with the growth of the Internet and its increasing social and political significance.

The rapid growth of the Internet; continuous change in Internet content and technology; and the Internet increasingly global character challenge the conventional regulatory approach to governance.

At all levels of government, policymakers have displayed an understandable inclination to rely on conventional regulatory tools to address governance challenges on the Internet. This regulatory approach – whether intended to discourage inappropriate Internet content or police undesirable behavior – has generally failed to achieve its goals. The Internet undermines regulatory governance by its very nature – its constant change renders out-of-date many regulations before they are fully promulgated; its international scope enables regulated parties who wish to avoid regulation to move off-shore easily; its enormous content, and effective anonymity, makes effective policing difficult.

The critical factor limiting the growth of the Internet today is the availability of affordable bandwidth. Increased bandwidth is necessary to keep up with current growth rates in Internet use and content. Greater bandwidth is also a necessary predicate for many of the latest Internet technologies and for the continued growth of e-commerce.

Competitors from telephone, cable, satellite and wireless industries are beginning to compete aggressively to meet the enormous demand for faster, more robust access to the Internet. These competitors – and others that may emerge – are deploying a variety of new broadband technologies to meet bandwidth demands in business and at home.

The extremely rapid expansion of access to the Internet should reassure some people who may once have had early concerns about the emergence of uneven access to the Internet. It is true that certain geographic areas of the country, primarily west of the Mississippi River, do not offer as much choice in access as in urban areas. But rhetoric about "digital divides" between information "haves" and "have nots" seems overstated in view of the fact that 78 percent of Americans have a choice of four or more service providers.

First, the demographics of Internet users more closely reflect those of America as a whole. Users do remain better educated and more affluent than the average American. But this is true for most early adopters of most new technologies, and even here, the differences between the Internet community and American society are diminishing.

On other factors, the "mainstreaming" of the Internet is more pronounced. For example, on metrics of race, sex, and age, the Internet community very nearly parallels American society, with most differences accounted for by other factors such as education and income. Early fears concerning demographically driven digital divides should also be assuaged.

Digital divides based on geography do remain, but they are relatively small, readily identifiable, and may be diminishing. Most Americans--a vast majority, in fact--enjoy several, relatively inexpensive, local alternative choices for dial-up access to the Internet.

A relative few have no local dial-up access to the Internet. For these Americans, located primarily in the West and South, the digital divide is real. For most others, basic, local access to the Internet is ubiquitous.

The Internet has achieved this extraordinary level of access as a result of growth and competition. Any government action aimed at redressing the small remaining digital divide should recognize this fact and carefully target government action to avoid disrupting or displacing this vigorous private market.