The earlier chapters of this report have chronicled changes in the connected population: who they are, where they are, what they are doing online, what devices and connection types they are using, and where they are using the Internet.  There is a sizable segment of the U.S. population (as of September 2001, 46.1 percent of persons and 49.5 percent of households), however, that does not use the Internet.  This chapter profiles this “unconnected” population and explores some of the reasons why it may not be online. 

The Offline Population

Table 8-1 at the end of this chapter presents the complement to Table 2-2—descriptive statistics for those individuals who do not use the Internet.  As the analysis in Chapter 2 shows, Internet use has expanded dramatically in the United States, but a number of groups are more likely not to be Internet users.  These non-users include:

·        People in households with low family incomes — 75.0 percent of people who live in households where income is less than $15,000 and 66.6 percent of those in households with incomes between $15,000 and $35,000.

·        Adults with low levels of overall education—60.2 percent of adults (age 25 +) with only a high school degree and 87.2 percent of adults with less than a high school education.[1]

·        Hispanics—68.4 percent of all Hispanics and 85.9 percent of Hispanic households where Spanish is the only language spoken.

·        Blacks—60.2 percent of Blacks.

Earlier chapters have examined the change in the online population focusing on the growth in the number of users or home connections.  We gain a different perspective by looking at the rate of decrease in the population that is not online.  In other words, we compare the change in the online population with the group initially not online instead of the group initially online.

 

Consider in the non-Internet-using population by educational attainment, for example.  Among people at least 25 years old with a high school education, the share not using the Internet declined from 69.4 percent in August 2000 to 60.2 in September 2001.  Over the same period and age level, the share of those with a college education who were not using the Internet shrank from 27.5 percent to 19.2 percent (Figure 8-1).  Thus, high school graduates had a slightly larger point change (9.2 percentage points) than college grads (8.3 percentage points).  Because so many more high school grads were not Internet users in August 2000, the 9.2 percentage point change over the next 13 months represented a 12 percent annual rate of decline in non-Internet-users (Table 8-2).  On the other hand, so few college grads were non-Internet-users in 2000 that their 8.2 percentage point change reflected a 28 percent annual rate of decline in non-Internet-users.  When Chapter 2 examined those same point changes from the perspective of the growth in Internet users, high school grads had a larger growth rate of Internet users than college grads (27 percent vs. 11 percent) (See Table 2-3).

 

Figure 8-1: Individuals Not Using the Internet, By Selected Educational Attainment Level,
August 2000 and September 2001

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

 

This observation applies to other comparisons.  For example, the group of non-users (25 years and older) with less than a high-school degree declined only 4 percent annually while those with a college degree—a much smaller group—dropped 27 percent on an annual basis (Table 8-1).  The percentage point differences vary similarly. 

The Importance of Cost to Households Never Connected to the Internet

The cost of Internet access matters much more to households with lower incomes than to those with higher incomes.  The September 2001 survey asked households without Internet subscriptions the question, “What is the main reason that you don’t have the Internet at home?”[2]  Survey results indicated that the largest specific response was that the cost was “too expensive.”[3]  This response was volunteered by one-fourth of these households, but much more often by lower income households than by higher income households.

 

Figure 8-2 shows the relationship between costs, income, and adoption of home Internet connections.  With successively higher income categories, fewer households report that cost is a barrier and more households are making their first connections to the Internet at home.  Households with incomes below $15,000 volunteered cost as the barrier to home Internet subscriptions 34.7 percent of the time.  Among households in that income category, the share of the population without home Internet subscriptions declined by only 6 percent between August 2000 and September 2001.  At the other end of the spectrum, only 9.6 percent of households with incomes of at least $75,000 said that they were deterred by cost.  That income level saw a 34 percent reduction in the share of households without home Internet between August 2000 and September 2001.

 

Figure 8-2- Adoption Rate and Internet “Too Expensive” by Income

Percent of U.S. Households without Internet

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

Among specific responses, cost rated highly across a number of demographic groups of non-Internet households.  In particular, respondents for married couples or single-parent families with children, and heads of households that were younger than 45 years of age, less educated, or unemployed all identified “too expensive” as the most important reason for non-connectivity at a much higher level than the national figure of 25.3 percent.

 

 

Why Households have Discontinued Internet Access

Those households that have discontinued Internet access numbered 3.6 million, or 3.3 percent, of all U.S. households as of September 2001.  Among this group of unconnected persons, cost was the most frequently cited reason for disconnecting (Figure 8-3).  Households with incomes less than $50,000 identified “too expensive” as the primary reason for discontinuing their Internet connection (26.9 percent of such households).  Cost was more important in households with only high school degrees (24.6 percent) than in households with college degrees (13.7 percent).  Those household heads younger than 45 rated cost (24.2 percent) more highly than household heads 45 years or older (19.0 percent).  Geographic differentials existed: households in rural areas cited cost less often (19.9 percent) than households in central cities (25.5 percent).

The lack of a computer or problems with the home computer also accounted for many persons discontinuing their use (Figure 8-3).  Although people have concerns about their children’s exposure to inappropriate material on the Internet (see discussion at the end of this chapter), this was seldom the reason cited by people who no longer subscribed to the Internet.

 

Figure 8-3 Reasons for U.S. Households Discontinuing Internet Access

 Percent Distribution, 2001

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

 

The Effect of Confidentiality Concerns

Some households may choose not to have a home Internet connection because of confidentiality concerns. The September 2001 survey asked respondents if they were more or less concerned about their confidentiality over the Internet as compared to the telephone.  It is important to note that although respondents reported being more concerned about their confidentially over the Internet, the question was phrased in such a way that respondents did not rate the degree of concern but rather whether they were more or less concerned.[4] 

About half (50.9 percent) of respondents were more concerned about their confidentiality over the Internet compared with the telephone.  About one-third (41.4 percent) of respondents reported their concerns were the same for both media, and 7.7 percent of respondents reported feeling less concern about confidentiality over the Internet compared to on a telephone.

In terms of age, persons under 25 years old were the least concerned about their confidentiality over the Internet (36.0 percent), compared with those 55 years of age or older who were the most concerned (54.8 percent).  A majority of respondents in the two age groups under 35 reported that they were either neutral or more concerned about the telephone.  In contrast, a majority of respondents over 35 were more concerned about the Internet than were either neutral or more concerned about the telephone.  Examining gender revealed that females and males shared a similar level of concern about confidentiality over the Internet:  51.8 percent compared to 50.1 percent, respectively.  Looking at household types, male-led households were least likely to be concerned about confidentiality over the Internet (41.4 percent), while female-led households were most concerned (54.9 percent).  Male-led households were also most likely to respond that there was no difference in confidentiality between the two media (52.5 percent), compared to 38.0 percent of female-led household who reported that there was no difference.

Content Concerns

Some households, particularly those with children under the age of 18, may choose not to have a home Internet connection because of the concern that the children may access inappropriate material.  As discussed in Chapter 5, the September 2001 survey found that among households with children, 68.3 percent responded that compared with material on television, they were more concerned about the kind of material children may be exposed to on the Internet.  This concern, however, did not translate into lower rates of Internet access among this group.  Among those that thought the Internet was a source of more concern than television, 51.8 percent had Internet in the home as compared to 48.2 percent who did not subscribe to the Internet.  Those who were less concerned (5.6 percent) or had similar concerns (26.1 percent) actually constituted a lower proportion of Internet households.

The Role of Network Effects

“Network effects” may be another factor determining whether people connect to the Internet.[5] Typically, the adoption of a technology that has “network effects” begins slowly. At some point a successful technology will reach a tipping point and adoption will accelerate rapidly.  As the technology saturates the market, the adoption rate slows, since most people who want the technology already have it.[6] 

Generally, the adoption of a technology does not take place uniformly across the entire economy or the entire population.  The penetration rate for fax machines, for example, is much higher among businesses than among households.  The fax never rivaled the telephone or mail for household communications, whereas businesses found considerable value in the near instant transmission of documents.  Under this concept, if a person’s family, friends, and broader community are Internet users, there would be increased incentive for them to go online.  On the other hand, if few of a person’s family, friends, or community were online, there would be less of an incentive to go online.  In looking at the relationship of home Internet to variables such as income, education, race or Hispanic origin, we may be picking up in part the probability of family, friends, and community to be online.

In sum, there are a number of reasons as to why a substantial proportion of U.S. households do not currently use the Internet.  Some reasons, such as those related to “network effects,” can be surmised from adoption patterns of new technologies.  Other reasons were provided as rationale in the September 2001 survey.  Cost (“too expensive”) rates highest among lower-income households, particularly for those that have decided to discontinue Internet access at home, and highly among many demographic groups of non-Internet households.  In contrast, confidentiality issues (even where households express greater concerns about the Internet than television) and concerns about how children use the Internet do not appear to be significant reasons why households stay or go offline. 

Table 8-1: Non-Internet Use From Any Location by Individuals Age 3 and Older,
October 1997, December 1998, August 2000, and September 2001

 

Table 8-2: Percent Difference and Growth Rates Non-Internet Use From Any Location by Individuals Age 3 and Older, October 1997, December 1998, August 2000, and September 2001