Chapter 4: How and Where America Goes Online

Internet users are expanding how and where they go online.  Faster connection speeds through digital subscriber lines (DSL) and cable are now available to more users as these technologies continue to expand their geographical reach.  New devices offer the opportunity for access without a computer and increased mobility of use.  Most striking, however, is the growth in the number of people who use the Internet from more than one location.

Connection Types: The Expansion of Broadband

Most individuals who use the Internet at home make that connection via a regular “dial up” telephone line (80.0 percent), with cable modems being the second most common way to connect (12.9 percent), followed by DSL (6.6 percent). 

Figure 4-1: Home Internet Connection Type, 2001
as a Percent of Individuals Using the Internet at Home







Source: NTIA and ESA, U.S. Department of Commerce, using U.S. Census Bureau Current Population Survey Supplements

Generally referred to as “broadband,” cable modems and DSL allow for higher speed access than is available through dial-up.[1]  The use of these services in residential settings has rapidly increased over the past year.  In August 2000, only 5.0 percent of all individuals or 11.2 percent of home Internet users claimed to have something faster than a dial-up service in their homes.  As of September 2001, those figures had risen to 10.8 percent of the population or 20.0 percent of individuals who use the Internet at home.[2]

This strong growth of approximately 116 percent over a 13-month period coincides with the growing availability of these services.  Until very recently, broadband was only available in selected areas of the country.  For example, the Federal Communications Commission (FCC) reported that high-speed subscribers were present in 75 percent of the nation's zip codes at the end of December 2000 as compared to 56 percent at the end of 1999.  The deployment of broadband occurred first in higher density areas.  According to the FCC, high-speed subscribers were present in 97 percent of the most densely populated zip codes at the end of December 2000 as compared to 45 percent of zip codes with the lowest population densities.[3]  As shown in Figure 4-2, differences by population density continued to carry over into 2001 with rural areas trailing urban areas and central cities.

Figure 4-2: Higher-Speed Internet Connection by Geographic Area
as a Percent of Total U.S. Internet Households

Source: NTIA and ESA, U.S. Department of Commerce, using U.S. Census Bureau Current Population Survey Supplements



The growth in broadband subscribership compares favorably to the deployment rates of other communications technologies and services.  Broadband deployment reached 8 percent of U.S. households in early 2001—an adoption speed that outstrips other technologies such as color television, cell phones, pagers, and VCRs.[4] 

Figure 4-3: Rate of Deployment of Selected Technologies

Source: eBrain Market Research and the National Cable and Telecommunications Association

Because cable and DSL Internet are more costly than dial-up services, the proportion of Internet users subscribing to these broadband services varies in expected ways, with individuals in high-income households, for example, having higher subscribership rates than individuals in lower income households.

Individuals with broadband access at home had a modestly greater likelihood of engaging in the activities shown in Figure 3-2.  For example, although 84 percent of Internet users send e-mail, that figure rises to 87 percent for broadband users specifically.  The only activity reflecting a large difference between broadband users and the Internet-using population, in general, is in the viewing of television or movies or listening to the radio.  In September 2001, 28.2 percent of broadband users engaged in these activities, compared to 18.8 percent of Internet users generally.

Spread of New Devices

The vast majority of Internet users in the United States still access the Internet through a desktop or laptop computer.[5]  Although the number of people using alternative Internet access devices is increasing, the survey revealed that people who use them typically also have a computer.  In September 2001, only 1.5 percent of the households that had home Internet access did not also have a computer.

The only category of alternative Internet access device owned by more than 2 percent of households is Internet-enabled cell phones or pagers (4.8 percent) and virtually all of these households also have computers.[6]  Only 1.8 percent of households include a household member who has an Internet accessible personal digital assistant (PDA) or other handheld device, and 0.6 percent of households have Internet access through a television-based system.  The television-based systems are the only category of alternative access devices where a substantial proportion of subscribing households do not also have a computer (44.4 percent), but this category accounts for 0.6 percent of total households.

Location of Use

Increased use of mobile Internet devices may eventually make the question of location less important.  For now, however, when most access still occurs through less than portable personal computers, where people use the Internet may have implications for the quality of access they enjoy (i.e., the degree of availability or access they actually have) or the type of activities they undertake online.[7]  For example, home Internet access may be thought of as a higher quality type of access because it is available (theoretically) 24 hours a day, seven days a week, while school or library access periods are limited to specific hours and often with time limits per session.

As shown in Figure 4-4, the most significant change between August 2000December 1998 and September 2001 is the substantial increase in the proportion of people who use the Internet both at home and from some other location.[8]  In 2000At the end of 1998, only 10.76.5 percent of the population used the Internet both at home and from some other location. In just under three yearsThirteen months later, that figure had almost quadrupled more than doubled to 24.5 percent.  That a growing number of people connect from multiple locations highlights the factcould indicate that the Internet is increasingly viewed as a basic communication and information tool, closer in nature to the telephone than the desktop computer.

Figure 4-4 also gives information as to the change in the proportion of the population using the Internet in each category: at home and from some other location.  In August 2000December 1998, 35.722.3 percent of the population used the Internet at home (25.015.8 percent plus 10.76.5 percent).  Home use had grown to 43.6 percent by September 2001.  Similarly, Internet use from a location outside the home grew from 19.417.0 percent to 34.8 percent over the same period.

Figure 4-4: Internet Use by Location
as a Percent of U.S. Population, 1998 and 2001

Source: NTIA and ESA, U.S. Department of Commerce, using U.S. Census Bureau Current Population Survey Supplements

“Outside the home” covers a variety of locations.  The September 2001 survey asks specifically about Internet use at six locations: work, school, public libraries, community centers, someone else’s house, and “somewhere else.”  Figure 4-5 shows that no single category of “outside the home” comes close to Internet use at home in terms of utilization by a proportion of the U.S. population.  However, percent of the total population may not be the most useful basis on which to consider location of use.  For example, the 51.9 million people who use the Internet at work represent 19.6 percent of the population, but 38.4 percent of those who work.  Similarly, the 31.5 million who use the Internet at school account for only 11.9 percent of the total population, but 44.8 percent who attend school.  Chapters 5 and 6 focus specifically on these school and work subgroups.

Figure 4-5: Internet Use by Specific Location
as a Percent of U.S. Population


Source: NTIA and ESA, U.S. Department of Commerce, using U.S. Census Bureau Current Population Survey Supplements

Of the universe of Internet users, 10.0 percent of Internet users access the Internet at a public library.  This proportion remained virtually constant between August 2000 and September 2001.  Over one-half of the population that uses the Internet at a public library is under age 25.  Additionally, 14.0 percent of Internet users do not use the Internet at home, school, or work. 

Internet use at public libraries varies by race and income:  only 8.6 percent of Whites that use the Internet use the public library as an access point, while the comparable figures for Blacks and Hispanics are 18.7 percent and 13.8 percent, respectively.  Among Asian Americans and Pacific Islanders, 11.6 percent of Internet users accessed the Internet at public libraries.

As shown in Figure 4-6, most of the people who use the Internet at public libraries also use the Internet at other locations.  Among racial and ethnic groups, 12.7 percent of Whites, 19.4 percent of Blacks, and 16.0 percent of Hispanics using the Internet at libraries do not also access the Internet from home, work or school.  Only 6.6 percent of Asian American and Pacific Islanders who use the Internet at a public library do not also use the Internet from some other location.


Figure 4-6: Public Library Internet Users, by Race and Sources of Other Access
as a Percent of Internet Users that Use Internet Facilities at Public Libraries



Source: NTIA and ESA, U.S. Department of Commerce, using U.S. Census Bureau Current Population Survey Supplements

Internet access at public libraries is more often used by those with lower incomes than those with higher incomes.  Just over 20 percent of Internet users with household family incomes of less than $15,000 a year use public libraries, and 6.1 percent of Internet users in this income category do not use the Internet at home, work, or school.  As household income rises, not only does the proportion of public library Internet users decline, but also the percentage of Internet users without alternative access points also declines.

Figure 4-7: Public Library Internet Users by Income and Location of Other Access
as a Percent of Internet Users

Source: NTIA and ESA, U.S. Department of Commerce, using U.S. Census Bureau Current Population Survey Supplements


[1] This study asked respondents about the two most common broadband technologies available in the United States, digital subscriber line (DSL) and cable modems plus any “other” higher-speed Internet access used at home.  These technologies usually feature broadband capabilities, although some applications or connections may possess speeds lower than the 200 kilobits per second—either in both directions or only upstream— that the Federal Communications Commission defines as “full broadband.”  See In the Matter of Local Competition and Broadband Reporting, Report and Order, CC Docket No.99-301 (rel. March 30, 2000) at ¶ 22.

[2] Homes with broadband have, on average, a higher number of individual Internet users.  For example, on a household basis, Internet connection through something other than dial-up increased from 10.7 percent to 19.1 percent between August 2000 and September 2001 among those households with home Internet connection—lower percentages than recorded on an individual basis.

[3] Http://

[4] Although it uses a different definition of “broadband” than is used in this report, Figure 4-3 provides a useful illustration of the relative deployment speeds of some familiar communications technologies.  See supra note 14 for discussion of broadband definition.

[5] This is not universally the case in other countries.  In Japan, for example, 30 million people access the Internet through NTT DoCoMo’s i-Mode using a handheld device.  Although iMode does not have full Internet capability, it is widely used in Japan to access the subset of Internet information available to subscribers.  See

[6] The questions on Internet access devices were asked only of the household. No information is available on actual use of these devices on a per person basis.

[7] The sample size of households that have Internet-enabled cell phones or pagers, but no computer, is too small for a reliable estimate to be reported.

[8] December 1998 is used as a basis of comparison rather than August 2000. The data from the August 2000 survey reflected the fact that students were generally not in school when the survey took place, and there appeared to be a downward bias on Internet use outside of the home.

Back To Table of Contents
Forward to Chapter 5