TeleCom Regulation - Some New Ideas

An article by Lars Poulsen

The Federal Communications Commission (FCC) is the US Federal agency writing and enforcing regulations of the Communications industries: Telephone, Radio, Television. For decades following the passage of the Communications Act of 1934, the field was stable enough that advances in technology could be accomodated with simple adjustments in regulations, but in the last decade, everything has exploded, as new media have sprung up, and old media have started merging. Congress is working on a major rewrite of the law, and the FCC is working on adjusting regulations to new technology. Here are some of the issues, but first a little background.

Background: A Brief History of the ATT Breakup 1985-1995

From about 1915 to 1985, telephone service in the United States was totally dominated by the American Telephone and Telegraph Company. For bookkeeping reasons, the company was organized into local telephone companies (generally one in each state), a long distance division, and an equipment manufacturing division (Western Electric). Each local operating company was regulated by state authorities, who limited to rates charged to subscribers, so that they would cover documented costs plus a modest guaranteed profit. By manipulating internal billing rates between divisions, profits (and taxes) could be shifted to whichever divisions the company wanted.

By the 1980's, a number of factors had lowered the cost of entry into the industry, and under the pressure of an anti-trust lawsuit presided over by Judge Harold Greene, A T & T agreed to split itself into about 8 new companies: Seven regional holding companies (with names including the words "Bell Telephone Company") which took over the existing local telephone service divisions, and a new conglomerate (which took over the ATT name) which absorbed the long lines division as well as Western Electric. After the assets had been divided between the 8 companies, the shareholders exchanged their old ATT stock for an equal number of shares in each of the new companies, which could then be traded separately. While the regional companies were still regulated monopolies, the breakup allowed competition in long distance telephone service, with provisions for each subscriber to separately select a long distance carrier.

The breakup had several important consequences:

All local telephone switches had to be equipped to handle the customer's selection of a long-distance carrier. This meant accellerating the replacement of many older electromechanical central office switches that were not yet worn out with new computer-controlled switches cabable of many other new services yet to be invented. Because these new switches were to be purchased by local telephone companies that were no longer a part of ATT, they were open to competitive purchasing without the former preference to ATT's Western Electric division. This created a greatly expanded market for both Western Electric and its competitors: Northern Telecom, Automatic Electric (which was by now owned by GTE) and several European manufacturers such as Siemens and Ericsson.
Under the old system, each state regulated the rates for local telephone service, and wanted to keep it as low as possible. This encouraged ATT to set rates for long-distance service higher than cost and use the revenue to subsidize local rates. With the breakup, long-distance rates had to be priced at rates that could compete with the new competitors: MCI and Sprint. Since nobody wanted to alienate the voters with sudden, massive shifts in rates, a system of settlement payments were set up to reduce the effects of the shift. Thus were born the "FCC mandated Subscriber Line Access Charge" that still appears on every monthly telephone bill, and the per-minute access charges that local telephone companies collect from the long-distance companies for each call.
In order to prevent the monopoly ATT from using its protected profits to enter new industries by shifting losses into the regulated cost base while keeping profits in unregulated subsidiaries, ATT had been specifically prohibited from entering into computer manufacturing and data processing. Since ATT was now no longer a monopoly, these bans were lifted. ATT immediately began selling computers to the competitive business and government market. After 5 years of continuous losses, they closed down the ATT computer division, but bought NCR (National Cash Register) to replace it, and after 5 more years of running this once successful supplier of computers to small businesses into the ground, they are just now shutting that down, too.
Unlike the old ATT, the new local telephone companies had no equipment manufacturing to protect, so they set much more liberal policies about connecting customer-owned equipment to the network; this spurred enormous growth in the manufacture of equipment such as small telephone switches (Private Branch Exchanges or PBXs), facsimile transmission devices (FAX) and modems.

Internet Access: Local or Long-Distance Service

When the network was divided into local services and long-distance services, judge Greene did not pay too much attention to a new industry that was just beginning: the online information service. Under the conditions before the breakup, such services were too expensive to attract many customers. A few years later, however, these had grown, and several of them had set up local access points in many cities across the US, and were even beginning to sell the transport of data between these access points as a separate service offering. From a certain perspective, this put them in competition with the long-distance telephone companies, which complained to the FCC and Judge Greene that the new data networks enjoyed an unfair advantage, since they did not have to pay the local network access charges of about 2 cents per minute on each end of the connection that ATT, MCI and Sprint had to pay to the local telephone companies. At the time, it was ruled that the emerging data network industry was a good thing which deserved a subsidy without which it might be unable to achieve critical mass, and besides, the amounts were too small to cause serious damage to the long-distance telephone companies. A regulatory notice was written which acknowledged the competitive similarity of longdistance telephone companies and data networks with local dial-in access, but granted the data networks an excemption from the local access charges.

The matter came up again in 1992 (?). By then it was clear that the data networks had taken hold, and the long-distance companies argued that the amounts were now significant. The FCC agreed, and drafted a Notice of Proposed Rule-Making to end the excemption. But they had overlooked one thing: As the data network had grown to significant volume, they had attracted large groups of intelligent, well-informed, well-connected users who deeply resented what they had dubbed "the proposed modem tax". A storm of mail to both the FCC and to the offices of federal legislators killed the proposal. For now.

With the explosive growth of the Internet in 1995, some enterprising network enthusiasts discovered that it was possible to get enough bandwidth between to PCs with high-speed modems across the Internet to transmit intelligible speech in real-time, thus allowing an Internet connection to substitute for a long-distance telephone call. This was a matter of great concern to many industries:

The long-distance telephone companies, as before, saw a loss of revenue to a competitor. They were less concerned than the last time around, though, because the shift in traffic would require increases in the bandwidth of the itnernal links in the Internet, which are generally leased from the same companies.
The local telephone companies, who were already complaining that the Internet users were taking advantage of "flat rate locall calling" to keep local telephone calls up for hours at a time instead of minutes. If they were now in addition to lose the access charges they were collecting for the long-distance calls, the financial damage would be serious.
The Internet service providers, who saw the potential for a vastly increased demand for bandwidth, for which they might not easily be able to recover higher subscription rates.

The local telephone companies have suggested to the FCC that it is time to end the excemption for local data network interconnections. But because of the nature of the Internet, the effects would be much worse than before: In fact, the effect could be an FCC-mandated surcharge of two cents per minute on every local call to a number answered by a modem, since the telephone company has no way of knowing which calls could pass data to the Internet.

In the opinion of this writer, this would be a terrible shift in policy at the worst possible time in telecommunications history, just before we are getting ready for competition in local access to provide a set of new options for access to the Internet. Here, then, are some possible outcomes:

Continued Residential Flat-Rate Local Calling

A large factor in the rapid growth of computer networks in the USA has been the tradition for a monthly subscription rate that includes unlimited local calls. This practice originally came about because it was too expensive to collect the data to charge for local calls. Tracking usage would cost more than providing the service. When technology for trackling and billing became less expensive, the regulators adopted a position that it was in the public interest to subsidize residential telephone service with revenue from business service, so that telephone service would be affordable to all households, even the poor, and that shifting some costs of residential telephone service to the business communitiy was fair, because the greater coverage of residences added value for the businesses would would recieve calls from the additional households. Thus, state regulators have tended to insist that residential services with flat-rate local calling be available everywhere, while allowing local telephone companies to offer an alternative service with metered billing for local calls, but with a lower monthly base rate. Business service, on the other hand, is (almost ?) universally metered, with rates around 2 cents pr minute during the business day and half that evenings, nights and week-ends.

The local telephone companies want to abolish flat-rate calling altogether. With increased residential modem usage, the load on the network arising from calls originating from residences is no longer negligible.

The internet service providers (ISPs) have mixed feelings about it: Without flat-rate calling, there would be much less interest in the network; but on the other hand, flat-rate tariffs encourage users to stay online longer, requiring the ISPs to install more phone lines and modems than would otherwise be needed for the same subscription revenue.

Mandated Feature Group B or D Trunks for Network Operators

If the proposed access charges go through, one likely form would be a brief statement that Internet Service Providers and other operators of "Value-Added Networks" (such as Compuserve, Telenet etc) would be required to connect to local central offices via Feature Group B or D trunks instead of regular business lines. This would be devastating to the Internet as we know it. Under such rules, my monthly bill for Internet access would go up from $25 to $175.

Feature group B trunks are the old way for long-distance carriers to connect to the telephone network, where the user dials a 7-digit number beginning with 950-xxxx. While this might not require any installation of new modems at the ISP office, it would require the ISP to pay the 2 cents per minute network access charge, while a caller with measured service would not be charged for these calls. To be economically viable, the ISP would have to charge an hourly rate for connect time of at least $2/hour. ($1.20 to the phone company, $0.80 to cover the administrative overheads of tracking and billing.)

Such a change would be good news for local telephone companies, long-distance telephone companies, centralized online information services (CompuServe, AOL). It would be bad for the users and the local ISPs.

The biggest problem with this regulatory tack is the definition of who is a network operator. Is a local BBS with no outside links a network operator ? (No. Clearly not.) If that BBS uses regular long-distance telephone calls to exchange electronic mail by modem with other BBS systems or with an Internet host, does this make the BBS a network opertor ? (Probably not.) If the BBS is connected to a BBS in another state via a leased line, but the line is only used to exchange email ? (Maybe not, but you see what is coming.) If the two BBS systems above are connected using Internet Protocols to each other, but not to the Internet ? (Probably yes!)

If a local business has several computers forming a local network with an Internet connection, does this require them to register as a network operator is they install a modem which an employee can dial in to from home ? Probably yes, or there will be no end to the arguing about when Internet access is material and when it is incidental.

All-measured Local Calling

As bad as it may sound, abolition of flat-rate local calling is actually better for the users than this worst-case scenario. Since the access fees are similar to the cost of measured local calls, and since both go to the local telephone company, it doesn't make a huge difference to the local telephone company, except that they get to discount them during off hours. It avoids the classification issue described above, allows the local ISPs to avoid collecting the charge, and does not set up the quandary of how to create a similar charge for a totally different technology when the cable companies start offering LAN service over their coax and fiber networks.

This solution is likely to emerge as a compromise, but it is still a large cost increase for people like myself. At the current per- minute charges for measured residential service, my $25/month bill will go up to $100/month.

A Better Idea

While we are at it, we should maybe re-think the whole relationship between local and long-distance telephpone service. If we are going to have competition on the local loop as well as in the long-distance network, it becomes impractical to manage the cross-subsidization from long-distance to local companies. I think it is time to suggest that a long-distance company be considered just another customer of the local telephone company, and be billed in the same way as any other business customer. Under this model, when I place a telephone call from Santa Barbara (California) to my wife's cousin in Orlando (Florida), there are three distinct segments to the call:

A local call from me to the long-distance company's local access point. This should be billed to me by the local telephone company according to the same rates I pay for other local calls.
A segment on the long-distance network. The long-distance carrier will bill me for this.
A local call from the long-distance company's switch in Orlando to my relative's house. The local telephone company will bill the long- distance company for this call; the long-distance company will obviously set its rates so that the cost of this segment is covered.

The long-distance company will want to get some services from the telephone company: First and foremost, they will want to receive the caller's billing number: Caller-ID, essentially. Of course, such services should be available on an equal basis to all other business subscribers. They will probably also want trunks with features that allow them to deliver a "fake number" to the recipients caller-ID display, so that the original caller's number can be presented instead of the number of the long-distance company's trunk. And indeed, I can see that other business subscribers may be interested in such a feature as well. (Before anyone claims that these two features are in contradiction of each other: I suspect that it would be in order for the number delivery interface to include a bit to indicate whether the number presented was the real billing number or not. I suspect that the long-distance company would be likely to reject all calls for which the billing number was suppressed.)

These changes may seem drastic to some; but I am convinced that they will simplify many of the decisions that must be made in the next few years as new services will make the network vastly more complex.