What Drives Broadband Adoption? An Aspen Institute Working Paper

WASHINGTON, December 5 – This report, authored by BroadbandCensus.com Editor Drew Clark for the Aspen Institute’s Communications and Society Program in August 2007, asks and attempts to answer a series of questions about broadband adoption.

Reports

Drew Clark, Editor, BroadbandCensus.com

Editor’s Note:This working paper was originally written for the Aspen Institute’s Communications and Society Program’s August 2007 forum in Aspen, Colorado, in which the author participated. At the time, the author was Senior Fellow and Project Manager at the Center for Public Integrity. The paper was included in “A Framework for a National Broadband Policy” (PDF). Republished with permission of the Aspen Institute.

What do broadband users want? The ability to connect online through some form of access, obviously. Service that doesn’t cost a fortune, clearly. Fundamentally and personally, however, what do broadband users want by going online? Why do 47 percent of adult Americans subscribe to broadband? Conversely, why do a little more than half not subscribe? Why do subscribers keep paying their monthly bills? In considering a framework for a national broadband policy, what can we learn from considering broadband adoption trends, both quantitatively and qualitatively?

In this paper, two specific questions about broadband adoption are addressed. Both are framed in the context of also considering the availability of broadband access and the affordability of available choices; those topics are explored in other papers. For this paper, consider:
• What other factors, such as equipment subsidies and consumer education, are necessary for encouraging adoption?
• What applications—such as telemedicine, e-government, or online education—are likely to increase demand for highspeed broadband access?

Both questions are viewed from the lens of the individual broadband user to determine why individuals subscribe, or fail to subscribe, to broadband. In the first section, some of the quantitative and qualitative research about broadband adoption are surveyed. In the second section, I offer my own set of questions and personal answers about the combination of applications, education, experience, and other motivations that lead an individual to subscribe. The next section offers tentative conclusions about the broadband applications on the “supply side.” And, the fourth and final section, offers tentative conclusions about some aspects of directed “education” and “subsidies” that could potentially stimulate demand.

What Do Researchers Say about Who Subscribes to Broadband?

Research on broadband adoption shows that Americans are adopting broadband. Put aside, for the moment, the debate about whether the United States is adopting broadband as fast as other developed nations— or developing nations. The Pew Internet and American Life Project’s annual and semiannual surveys about broadband adoption show a consistent pattern of increase. Figure 1, from the June 2007 Home Broadband Adoption report, by John Horrigan, Associate Director for Research, and Aaron Smith, Research Specialist, shows the breakdown of broadband adoption across various demographic categories.

Figure 1: Trends in Broadband Adoption Across Population Subgroups (see PDF above for figures and tables)

Pew’s 2005 report argued that broadband adoption at home in the U.S. was “growing but slowing.” The 2005 report created the following model of broadband adoption:

• People do more things online the longer they’ve been online.
• Dial-up users are more likely to want broadband the longer they’ve been online.
• Not everyone wants broadband—and the people who do not want broadband typically have less online experience and are processing fewer bits.
• High-speed users switch to broadband to processmore bits, less so because of price.

Under this model, the decision to get broadband depends on the “intensity of Internet use,” which in turn is a function of time online and connection speed.69 Considering this model, Horrigan concluded in 2005 that although “years of online experience” may have driven broadband adoption in 2002, early in the growth phase, that was no longer the case in 2005.

On the one hand, this is not too surprising—early adopters, the “low hanging fruit,” have been picked. But it is important to recognize that there could be very different migratory patterns toward broadband. Internet use, rather than tapering off in recent years, could have continued its late ’90’s-early 00’s upward climb. Broadband prices could have been on the decline or network speeds might have improved substantially. That or other forces might have meant more switching from dial-up to high-speed and more adoption “de novo” of high-speed by new users.

Somewhat unexpectedly, the Pew 2006 report found home broadband adoption growing 40 percent from March 2005 to March 2006— twice the growth rate of the preceding year. Horrigan writes, “A significant part of the increase is tied to Internet newcomers who have bypassed dial-up connections and gone straight to high-speed connections. This is a striking change from the previous pattern of broadband adoption.” Among the factors, many of them new for that year, Horrigan identified:

• There was strong growth in broadband adoption by African Americans and by people with low levels of education.
• Digital subscriber line (DSL) market share increased, driven by aggressive price-cutting by DSL providers.
• About 48 million Internet users were posting online content, the majority of whom are home broadband users.
• Awareness about Voice over Internet Protocol (VoIP) increased 86 percent between February 2004 and December 2005.

Jump forward one more year, to the June 2007 report, and the adoption growth rate is down again. Figure 2 is Pew’s chart of year-to-year growth grates in home broadband adoption.

Federal Communications Commission (FCC) and Pew data from 2003 to 2007 show similar trends year-to-year growth grates in home broadband adoption. The number of “high-speed lines” (200 kbps in either direction) grew 32 percent, from 32.5 million on June 30, 2004, to 42.9 million on June 30, 2005. The number of such lines grew 52 percent, to 64.6 million, by June 30, 2006.

Figure 2:Year-to-year growth rates in home broadband adoption

Of those 64.6million lines (themost recent total fromthe FCC), 50.3 million served primarily residential end users. Of those residential broadband connections, the FCC reported that 55.2 percent of subscriptions were cable modem connections, 40.1 percent were asymmetric DSL connections, 0.2 percent were symmetric DSL or traditional wireline connections, 0.9 percent were fiber connections, and 3.7 percent were other types of technologies, including satellite, terrestrial fixed or mobile wireless (licensed or unlicensed), and electric power lines. The FCC says that broadband is available via DSL to 79 percent of local telephone company subscribers and via cable modem to 93 percent of cable television subscribers.

It is increasingly clear that there are two major groups of people who have not yet subscribed to broadband: dial-up users and non-Internet users.Dial-up usersmay be “happy dial-up users” because they get what they want out of their slower Internet experience. Alternatively, they may be frustrated dial-up users because of price or, more likely, availability constraints on broadband.

Non-Internet users have rejected the Internet experience, for whatever reason. Occasionally, as is evident in the spike of broadband adoption from March 2005 to March 2006, they can be lured directly to broadband subscriber status. Many, however, simply wish to avoid aspects of the Internet, such as pornography and the threat of various forms of identity theft.

Pew also has survey results on some of the reasons that individuals choose to take broadband, based on three separate surveys—January 2002, February 2004, and December 2005 (Table 1).

Table 1: Reasons for choosing high-speed Internet connection at home

The leading response to the survey: “Faster access/Greater speed” in January 2002 and December 2005 and “Previous connection was too slow/frustrating” in February 2004. The latter response may be effectively identical to the former. Indeed, at the spring 2007 meeting of the Aspen Institute Roundtable on Spectrum Policy, Andrew McLaughlin of Google gave a great definition of broadband: when a user isn’t constantly frustrated with the Internet experience.

If the goal is to get more people to subscribe to broadband, exclusive of considerations of availability and price, then happy dial-up users and non-Internet users are key groups to be targeted.

What Do Individuals Say aboutWhy They Subscribe to Broadband?

Understanding the demographics of broadband subscription begins to put some substance behind our key inquiry: How can individuals be motivated to subscribe to broadband?

The following model may be useful for thinking about this question:

1. Think of yourself: When did you subscribe to broadband in your home, and what led you to subscribe?
2. Think of other Americans, particularly the “happy dial-up users” and the “Internet rejecters.” How would a pitch to subscribe to broadband be targeted at them?
3. Think about individuals facing the prospect of adopting broadband in other parts of the world, such as China. It may be more exciting to consider a “fresher” market than the United States, with its more mature stage of broadband adoption.

Some questions to ask include:

• How long did you use the Internet before subscribing to broadband at home?
• How frequently did you experience the broadband Internet (i.e., at university, in the workplace) before subscribing at home?
• What companies were offering service to your home, what type of service were they offering, and at what price?
• What applications tipped the balance in favor of your subscribing to broadband at home?
• Were any other factors involved in your decision to subscribe to broadband at home?

Here are my own answers:

• I used a primitive form of Internet access, via an America Online dial-up connection, through a creaky Apple MacIntosh in February 1995. I first saw the high-speed Internet at Columbia University in August 1995. I finally subscribed to broadband on March 14, 2004—making nine years of Internet use before subscribing to broadband.
• I used broadband constantly at school, and then at work, in the years since 1995. My extensive use of broadband and work probably was a major factor in delaying my personal broadband adoption.
• I did not inquire about broadband availability in the homes and apartments I moved into in 1996 and 1997.When I moved to a home in 1999, I didmake an inquiry about DSL broadband availability (it was available), but I did not subscribe. When I decided to subscribe, I tried DSL, but the service did not work; I then subscribed to cablemodemservice. (I believe the price of DSL was $40,when included with traditional phone service; the price of cable modem service was $40, when included with basic cable television.)
• Saving money by subscribing to Voice over Internet Protocol (VoIP) service was the primary deciding factor in my decision to subscribe to broadband. I cancelled local telephone service and Internet service. A second motivating factor was the ability to get basic cable television programming—that is, assembling an ad-hoc “bundle.”
• A final factor motivating adoption was simple embarrassment: How could I be a decent technology journalist and not subscribe to broadband at home?

My responses offer one personal window on broadband adoption. I have asked the same questions of friends, neighbors, colleagues, and sources. I’d like to see and participate in ways to publish more of these responses. This kind of qualitative, even anecdote-driven, research also is instrumental in helping us better understand broadband deployment. Indeed, when I interviewed John Horrigan about this subject, I asked the same questions of him. He told me that he made the transition from dial-up to broadband in 2003 and that one of the factors influencing the decision was that his employer agreed to pay for a home broadband subscription. Cisco Systems is another company that pays the home broadband subscription costs as an employee benefit.

Broadband Applications on the “Supply Side”

Although speed frequently is identified as the reason for broadband subscription, my personal experience suggests that usually some particular application (or combination of applications) causes an individual to reach the tipping point. In my case, it was VoIP. Almost immediately thereafter, I installed a WiFi router, enabling broadband access anywhere in the house. That technology, in turn, facilitates a host of additional applications, any one of which could be the tipping point for others to subscribe to broadband.

Other heavily used high-bandwidth broadband applications in the Clark home include the following:

• Google Earth (Three-year-olds and seven-year-olds love it!)
• Educational videos and games
• Blogging
• Smugmug photo-sharing
• Video and audio streaming, including Internet radio
• Google Calendar for sharing schedules
• “Presence,” in the formof G-mail/instantmessage integration, etc.
• Online classes.

An application such as VoIP can prove successful in motivating a broadband purchase because it takes broadband off the desktop/laptop and into another device—such as a telephone—that is frequently used. I have been disappointed that equipment manufacturers and webcasters have not taken better advantage of opportunities to embed Internet radio applications into dedicated, IP-centric devices. Of course, the TV-PC convergence remains, after all these years, very much a work in progress. When I was watching an important cablecast that began to experience technical difficulties, I fired up my laptop and watched the webcast version of the program. Viewing on the larger TV screen was not possible, however.

In addition to IP-centric capabilities taking over telephones, radios, and televisions, such capabilities integrated into refrigerators, freestanding Webcams (whether for security or other purposes), or other household devices may reach those “happy dial-up users” and even some Internet rejecters. It is better to think about such applications in specific rather than general terms. In other words, diabetes patients or prospective diabetes patients may be motivated to subscribe to broadband to participate in a specific experimental trial but not to take advantage of “telemedicine” in general. The ability to enroll in a specific class may motivate a broadband purchase. The ability to do a job from home and avoid a commute is likely to be another key motivator in nudging broadband subscriptions upward.

Educating and Subsidizing for Broadband Demand

What forms of subsidization and education are necessary to stimulate demand for broadband? In the case of subsidization, consider various potential subsidizers: governments, employers, access providers, educators, and advertisers.

Subsidization of Internet services by the government or a business partner interested in advertising is central to many municipal wireless build-outs, including services to be offered by Earthlink in Philadelphia. In San Francisco, Google will subsidize a slower, ad-sponsored version of the wireless service. Other nationwide proposals, including that of M2Z Networks, contemplate free nationwide wireless Internet access through a 20 megahertz block of radio frequencies. As discussed above, employers play an important role—possibly a crucial role—in subsidizing their employees’ broadband use to facilitate work from home. According to a study by RVA Market Research for the Fiber to the Home Council, 13 percent of home fiber optic users work from home more often—a monthly average of 7.3 more workdays at home instead of the office. In most of these cases, having a fiber-optic connectionmade their employers’ attitude toward teleworkmore favorable.

Nevertheless, most discussions about subsidization deal with the Universal Service Fund’s (USF) system of cross-subsidization to broadband services offered by carriers, not subsidization of services or goods purchased by a consumer.

Equipment subsidies have received even less discussion. Here the question must be:What device to subsidize? Among the choices are the following:

• WiFi or other wireless-enabled laptops
• WiFi routers
• Wireless access devices (for non-WiFi fixed wireless services, such as a satellite dish in a rural area)
• Other standalone health- or home security-related IP devices.

Ironically, Congress has not chosen to subsidize any IP device at all. Instead, it has chosen to offer a $40 subsidy for a converter box that allows an analog device to receive digital television broadcasts. Aside from television, subsidies for equipment seem like a stretch for the government and for employers, for the simple reasons that prices are always dropping and government always seems to have more pressing priorities for its funds. Finally, worth noting is the fact that access providers routinely subsidize equipment (e.g., cable modems and wireless access devices) as part of a package of paid Internet service.

A final point for consideration is what kind of education consumers need to understand their broadband options. “Education” can include basics such as computer and Internet literacy. In most cases, this basic education is a prerequisite for home broadband use. Education also can include broader information about the true availability of broadband services in one’s area—as well as information about actual offers of service. The Center for Public Integrity’s Well Connected Project is engaged in one aspect of this effort: seeking to publicly display the names of each company that provides broadband within a particular ZIP code. If this effort is successful, it could enable consumers to see a complete list of all companies that offer broadband within their geographic area. The Federal Trade Commission (FTC) also intends to monitor the information that telecommunications and cable companies provide about high-speed Internet service in the service plans they offer to customers.

Broadband Breakfast Club:

Editor’s Note: Join the next Broadband Breakfast Club on Tuesday, December 9, 2008, on how broadband applications – including telemedicine – can harness demand for high-speed internet services. Register at http://broadbandbreakfast.eventbrite.com