The Internet Access Coalition

Before this domain compared providers, it was the public face of a 24-member industry coalition that spent 1996 and 1997 arguing that the dial-up internet should not be billed by the minute.

A tax on the dial-up internet

In 1996 the American internet ran on top of the voice telephone network. A home computer dialed a local number, a modem negotiated a connection at 28.8 or 33.6 kilobits per second, and the call stayed open for as long as the session lasted. The subscriber paid an internet service provider a flat monthly fee, commonly around 19 dollars, and paid the telephone company nothing extra, because the call itself was local. That arrangement was not simply how the market had settled. It rested on a regulatory classification, and in mid-1996 the largest local telephone companies asked the Federal Communications Commission to change it.

The request was specific. Local carriers wanted permission to assess interstate access charges, billed by the minute, on the providers that answered modem calls. Access charges were the tolls long distance carriers paid for the use of local networks at each end of a call, and they had never applied to internet providers. Applied now, they would have flowed straight through to subscribers. The coalition's own estimate was up to 60 cents an hour, which on a modest 20 hours of monthly use worked out to about 12 dollars added to a 19 dollar bill. Flat-rate pricing, the thing that made unlimited browsing thinkable, would not have survived that.

Why two dozen firms joined forces

The Internet Access Coalition was assembled in response, and its roster shows how broadly the threat was read. This was not a narrow lobby of internet service providers defending a revenue line. It was most of the computer industry. Sixteen companies signed on: Apple Computer, Compaq, Dell, Digital Equipment and IBM from the hardware side, Intel and Sun Microsystems from silicon and systems, Microsoft, Netscape, Novell and Oracle from software, America Online and CompuServe from the online services, EarthLink as a pure dial-up provider, and Eastman Kodak and GE Information Services from further afield.

Eight trade associations supplied the Washington machinery: the American Electronics Association, the Business Software Alliance, the Consumer Electronics Manufacturers Association, the Information Technology Association of America, the Information Technology Industry Council, the Software Publishers Association, the Internet Service Providers and Users Association and the Voice on the Net Coalition. Twenty-four names in total, and they had almost nothing in common commercially. Netscape and Microsoft were a year into the browser war. Apple and Compaq were fighting over the same shelf space. What united them was that each sold something whose value depended on ordinary people staying online without watching a meter.

"The affordability and usefulness of the Internet are in jeopardy."

Internet Access Coalition, Keeping The Internet Affordable, 1997

A single-issue group with a website

The coalition was light infrastructure by design. A steering committee ran it, chaired by Paul Misener, who also served as the public contact for policy questions. Press work went through Angela Leung of The Dittus Group, a Washington communications firm. There was no membership program and no conference. There was a docket, a study and a website at internetaccess.org.

The site was small and it did specific jobs. An introduction laid out the argument in language a non-specialist could follow. A page on the FCC proceeding explained what the Commission had actually proposed and how to file a comment on it, down to the docket number that had to appear in the subject line. A members page listed the roster, a take-action page told readers what to do next, and a study section carried the coalition's economic evidence in full. Page footers read Copyright 1997 Internet Access Coalition. The underlying template dates to January 30, 1996, and the surviving pages carry revision stamps from February and March 1997, which is when the campaign was at full stretch.

The engineering concern was real

It is worth being fair to the telephone companies, because the version of this story where they were simply grasping is the less interesting one. The public switched telephone network was engineered around a statistical assumption: that a residential line would be in use for a few minutes at a time, that most lines would be idle at any given moment, and that shared equipment could therefore be sized well below the number of lines it served. A modem session that held a line for an hour, or three, violated that assumption. Long holding times genuinely do load a local switch differently from voice calls, and the carriers were not inventing the phenomenon.

Between June and October 1996 five documents made that case in the record. Bell Atlantic filed a report on June 28. Pacific Bell followed with an ESP impact study on July 2. NYNEX submitted a letter on July 10. U S West filed an ESP network study on October 1. Bellcore, the research arm the regional companies jointly owned, contributed an analysis by Amir Atai and James Gordon. The rhetoric ran ahead of the filings: Pacific Telesis chief executive Philip J. Quigley told the San Francisco Examiner on October 4, 1996 that the network faced a meltdown.

"the sole grocer in town demanding payment today for a pie made in a bakery that the grocer might build in the next decade"

Internet Access Coalition, on paying for services not yet built, 1997

Answer it with measurement

The leap from a real phenomenon to a per-minute charge on an entire industry required showing three things: that the congestion was widespread rather than local, that it was expensive, and that the carriers were not already being paid for it. Those are empirical questions, and the coalition chose to treat them as such rather than as a debate over principle. It commissioned an independent report from Economics and Technology, Inc., a Boston consultancy that had spent two decades doing telecommunications cost analysis for regulators and intervenors.

ETI's Lee L. Selwyn and Joseph W. Laszlo delivered The Effect of Internet Use on the Nation's Telephone Network on January 22, 1997. It ran past fifty pages across five chapters, worked through each of the five carrier filings in turn, and reached three conclusions: that data traffic posed no significant threat to network integrity, that the revenue the carriers were already earning from internet-driven demand exceeded their own estimate of what reinforcement would cost by roughly a factor of six, and that the durable answer to growing data traffic was data-friendly network investment and local competition rather than usage charges on a voice network.

Making money, not losing it

The economic half of the coalition's case was the part the carriers found hardest to answer. Every dial-up call was already a paid call under state tariffs, billed to or bundled for the calling party. More to the point, the internet was selling telephone lines. Additional-line penetration in American households had run at roughly 2.5 to 3 percent before 1990 and reached 14.7 percent by 1995, and by that year about six million residential lines were being used primarily for online access. ETI put the cumulative revenue from second residential lines between 1990 and 1995 at 3.5 billion dollars, most of it earned on copper that was already in the ground.

The coalition's framing of the deal on offer was less technical. The carriers were asking to be paid now for data-friendly services they had not yet built and had not committed to building. The introduction page made the comparison that has outlived every other sentence on the site, and the answer it implied was not that the phone companies should get nothing, but that they should build something first and then earn a return on it the way firms in competitive markets do.

What the coalition won

The FCC's access reform proceeding, opened with a notice of proposed rulemaking on December 24, 1996, tentatively concluded that internet access providers should not pay interstate access charges. The coalition's job from that point was to keep a tentative conclusion from being reversed, which it did through comment filings, through the study, and through pushing ordinary users to write in. The exemption held. Flat-rate internet access remained legal, dial-up subscriptions kept climbing, and the per-minute internet never arrived in the United States.

The companion half of the proceeding mattered nearly as much. A notice of inquiry issued the same day asked what it would take to get faster, data-friendly connections built, with comments due March 24, 1997. The coalition's answer was ADSL and HDSL, technologies that already existed and that it argued could raise connection speeds by a factor of a hundred over the 28.8 and 33.6 kilobit modems of the day. That is a reasonable description of the DSL rollout that followed, and of the broadband market that eventually replaced dial-up entirely.

A coalition with a finite life

Single-issue coalitions are built to dissolve, and this one behaved accordingly. Once the exemption was safe there was no second campaign and no rebranding into a permanent trade body. The website simply stopped changing. Captures show the coalition pages standing largely as they were into 2001, with a short-lived membership-login experiment appearing near the end.

What the coalition left behind is a small, unusually clear artifact of how internet policy was argued before internet policy had a permanent constituency. The trade groups it worked through still exist. Roughly half the companies on the roster have since been acquired, merged or wound down, which is part of why the list is worth preserving: it is a photograph of who had a stake in an open, flat-rate internet in early 1997, taken a year before Google was founded. The pages are preserved here at their original addresses.