| In the Matter of | ||
| Access Charge Reform | CC Docket No. 96-262 | |
| Price Cap Performance Review for Local Exchange Carriers | CC Docket No. 94-1 | |
| Transport Rate Structure and Pricing | CC Docket No. 91-213 |
|
Scott J. Rubin, Esq.
3 Lost Creek Drive
Selinsgrove, PA 17870-9357
(717) 743-2233
sjrubin@ptd.net
Counsel for
Pennsylvania Internet Service Providers
Dated: January 27, 1997
Background and Need for Change
Fifteen Internet service providers (ISPs) in Pennsylvania are jointly submitting these Comments to urge the Commission to refrain from taking action that would harm the competitiveness of small ISPs. These 15 ISPs serve primarily small cities and rural areas and, at the present time, are captive customers of their LECs for most telecommunications services. These ISPs' particular concerns can be summarized as follows.
Changing the cap on the subscriber line charge ("SLC"). The Commission should not increase the current cap on the SLC. Increased competition in the local loop, coupled with continued dramatic increases in interLATA calling minutes per line, will work in tandem to reduce the local loop costs per line and reduce the per-minute access rate. Any increase in the SLC will have an adverse effect on the ability of people, particularly in small cities and rural areas, to access the Internet. In addition, there is no compelling reason to permit the SLC to be geographically deaveraged. The available data show that customers in rural areas are already paying more in total access charges per line than are customers in urban areas. There is no reason to further handicap the ability of those in smaller communities to gain access to the telecommunications network. This discussion begins on page 8.
Determining the SLC for ISDN lines. The SLC for primary rate interface ("PRI") ISDN lines should be set at two times the SLC for an analog line. This approach is supported by the physical nature of the facilities and is not inconsistent with the likely cost of providing PRI ISDN service in the future. This discussion begins on page 15.
Removing high-capacity services from price cap regulation. High-capacity services must not be removed from price cap regulation until real competition exists in a central office for providing these services. Many smaller businesses, including ISPs, are relying more heavily on high-capacity services. Unless a customer actually has multiple competitive options available to it, the price of such services should continue to be regulated. This discussion begins on page 17.
Regulation of terminating access. The Commission must not impose any fee on consumers for receiving telephone calls. While it might be appropriate to reallocate charges to IXCs for terminating and originating access, substantial harm would result to all types of consumers, including ISPs, if end users are actually charged for the receipt of telephone calls. This discussion begins on page 19.
Treatment of interstate information services. The Commission has consistently held that information service providers, including ISPs, are end users of telecommunications services; they are not providers of such services. This holding should not be changed. When the nature of the services provided by ISPs is examined, it is clear that they are substantial end users of the local communications network; they are not telecommunications service providers. As such, they should continue to pay the SLC and they should not be subject to access charges. This discussion begins on page 21.
The impact of proposed changes on small businesses. The Commission should be aware that several aspects of its decision, primarily those mentioned above, could have a serious impact on small businesses. If the Commission does impose substantial additional costs on small businesses, it must prepare an analysis under the Regulatory Flexibility Act to evaluate the impacts of such changes. This discussion begins on page 26.
Pursuant to the Notice of Proposed Rulemaking ("NPRM") issued in these dockets by the Federal Communications Commission ("FCC" or "Commission") on December 24, 1996, Pennsylvania Internet Service Providers ("PaISP") submits these Comments for the Commission's consideration.
PaISP is an ad hoc group of 15 small, independent Internet service providers ("ISPs") that provide service throughout the Commonwealth of Pennsylvania. Figure 1 shows in gray the portions of Pennsylvania that are served by one of more of these ISPs. These 15 ISPs have joined together in order to have the Commission consider their concerns in deciding whether to amend the regulations concerning access charges and other matters. The ISPs that make up PaISP are listed in Appendix A to these Comments. These Comments represent the consensus views of the 15 ISPs. For ease of reference PaISP will be used in the singular as a shorthand way of referring to all 15 companies collectively.
The 15 ISPs participating in these Comments, collectively provide more than 20,000 customers, most of whom are residences and small businesses, with Internet access in all portions of Pennsylvania. They include among their customers approximately 200 schools, 40 hospitals, 60 libraries, and 70 local governments. Most of these customers are in small cities and rural areas that would not have toll-free access to the Internet without a small, local ISP. In order to bring the Internet to these communities, these ISPs purchase more than 2,500 local access lines from their local exchange carriers ("LEC") - either Bell Atlantic-Pennsylvania or GTE North. Annually, these ISPs collectively pay more than $1.5 million to their LECs, representing anywhere from roughly one-eighth to nearly one-half of each ISP's total revenue.
The companies in PaISP vary in size - from 25 access lines to 500 access lines - but all are small businesses by any definition. But they also are substantial users of telecommunications services who must purchase essentially all of those services from their LEC. Many of these ISPs are located in small cities or towns that do not now, and will not for the foreseeable future, have any access to competitive providers of the telecommunications services that they need - analog and ISDN local telephone lines; fiber optic facilities; T-1, SMDS, and other high-speed data lines; and similar facilities.
Generally, PaISP is concerned that any significant increase in the charges that are imposed to recover the interstate portion of access costs would have a serious and detrimental impact on small ISPs. Each of these companies is a small business, typically providing less than a few thousand customers - nearly all of whom are residential - with access to the Internet. These are small, independent businesses that typically invest nearly all of their excess revenue in new plant and equipment, either to improve service to existing customers or to expand their service area. Moreover, nearly all of these ISPs have been in business for less than two years and, thus, are still attempting to recoup their start-up expenses and otherwise keep their fledgling businesses profitable.
The market for Internet access is extremely competitive today. Indeed, many of the ISPs that have joined in these comments compete directly with each other. In that market, residential customers are demanding, and are receiving, unlimited Internet access for a flat rate (typically about $20.00 per month). At that price level, every dollar that a small ISP has to pay to its LEC is a dollar that is not available to install new equipment or otherwise expand and improve the quality of its service.
As it stands today, these companies typically pay anywhere from one-eighth to nearly one-half of their revenue to their LEC in order to purchase the physical connections that are necessary to provide Internet access. It is unlikely that these businesses can withstand an increase in that ratio and continue to invest in new equipment. So, if the charges that they must pay to their LECs increase significantly, there are only two likely results: either small ISPs will go out of business or they will stop expanding their networks (which probably will lead to their going out of business in the future). A third option, that small ISPs will increase their prices, is highly unlikely. As long as there are larger ISPs (including the ISPs that are affiliated with LECs) that are willing to provide flat-rate service, it will be extremely difficult for a small ISP to charge on any other basis.
Simply, PaISP is made up of 15 small businesses - many of them in small cities and towns - that are relatively large users and purchasers of telecommunications services. As this Commission has found consistently, ISPs as enhanced service providers are customers of the LECs; they are not providers of communications services themselves. See, e.g., 47 CFR § 69.2(m) and In the Matter of Amendments of Part 69 of the Commission's Rules Relating to Enhanced Service Providers, 3 FCC Rcd 2631 (1988). Indeed, in some exchanges, an ISP is one of the largest customers of the LEC.
PaISP will be focusing its specific comments in this proceeding on the ways in which this Commission can ensure that it is not tilting the playing field in such a way that small ISPs will be harmed. Where LEC-specific examples are provided, they will be provided for LECs in Pennsylvania. PaISP believes that similar concerns exist for small ISPs nationwide.
The end-user or subscriber line charge ("SLC") was originally established to enable LECs to recover the cost of the local loop that is allocated to the interstate jurisdiction. When it was first adopted, these costs were to be recovered through two separate charges to the end user: a charge on each access line and a charge on each minute of use. In the Matter of MTS and WATS Market Structure, 93 FCC2d 241 (1983), paragraphs 176-194. Importantly, in an era of cost-based rates, each customer would have been required to pay a minimum charge (the per-line charge) and would have been subject to a cap on the total charges that it would be required to pay. This limit on the maximum charge would serve to protect each customer from paying more than its fair share of the interstate portion of the local loop costs. The Commission then believed that this would help consumers make informed choices among alternative methods of obtaining interstate access.
However, this cap on interstate charges was soon abandoned. In its first reconsideration order, the Commission decided that it would simplify the implementation of these new charges to eliminate the end user usage charge, increase the end user line charge, and place more of the costs on the carrier line charge. In the Matter of MTS and WATS Market Structure, 97 FCC2d 682 (1983), paragraphs 26-33. As the Commission noted at that time, the major detriment to this new approach was the elimination of the cap. Consequently, a customer who had a high volume of interLATA calls could end up paying substantially more than the cost of providing that access.
Needless to say, a great deal has changed in the 14 years since those initial orders were issued. Price caps have replaced rate of return regulation, making it much more difficult to track the actual cost of providing interstate access service. The number of access lines per person and the volume of interstate calling per line have both increased dramatically. The costs of technology have fallen dramatically. In short, we have witnessed a revolution in telecommunications - increased competition, new services, falling costs, and an explosion in the way in which people use the telecommunications network.
A few examples of the changes between 1984 and 1997:
These types of changes should send a clear message to policy makers: something is working; drastic changes should not be made. Costs are falling; competitive options are increasing; the Internet is growing at explosive rates; consumer use of the telecommunications network is growing rapidly. PaISP respectfully suggests that in this type of environment, regulators should proceed with great caution.
To be sure, there are changes that need to be made to encourage further competition within the local loop. Such competition will serve to benefit all users of telecommunications services in general, and larger-volume users like ISPs in particular. But the Commission must be careful that it does not make a mistake that could have the unintended effect of causing a serious disruption in these trends.
In paragraphs 65-67, the Commission seeks comment on three important issues: (1) whether to increase or remove the cap on the Subscriber Line Charge ("SLC"), (2) whether any such increase should be phased in, and (3) whether the SLC should vary by geographic region within a LEC. Each of these issues is critically important to PaISP.
Any increase in the SLC would have two serious effects on small ISPs and their customers. First, it would increase the ISPs' costs of doing business which may result in increased costs to the consumer, an inability of ISPs to continue to expand and improve their service, and additional obstacles in the expansion of Internet service into rural areas. Second, it would directly increase the cost to the consumer. Many consumers are installing second telephone lines in order to facilitate access to the Internet. The cost of such an installation would increase significantly under some of the proposals being considered by the Commission, particularly those that would eliminate the cap on the SLC for all but the first residential telephone line.
Stated differently: Why would the FCC do anything that would be likely to reduce consumers' access to the Internet? Yet that is precisely what increases in the SLC would do - it would "raise the bar" for a residential user to access the Internet and it would restrict the ability of ISPs to expand their service.
If such an increase were required in order to foster competition in the local loop, then it might be worth taking the risk. But it is not necessary to increase the SLC in order to bring about more competition for the provision of local exchange service. There are several facts that support this view:
|
|
|
|
| |||||
| Number of Access Lines | ||||||||
| Business single line | 14,800,956 | 14,718,031 | 15,156,744 | 12,488,621 | 12,167,769 | -17.8% | ||
| Business multi-line & digital | 23,632,737 | 24,017,471 | 25,574,751 | 29,917,748 | 33,421,889 | 41.4% | ||
| Residential | 90,836,051 | 93,235,927 | 95,599,309 | 98,224,590 | 101,333,305 | 11.6% | ||
| Total | 129,269,744 | 131,971,429 | 136,330,804 | 140,630,959 | 146,922,963 | 13.7% | ||
| Total InterLATA Billed | ||||||||
| Access Minutes (x 1000) | 405,456,046 | 432,356,515 | 465,270,370 | 500,297,266 | 549,982,267 | 35.6% | ||
| Calculated Ratios | ||||||||
| Billed access minutes per line | 3,137 | 3,276 | 3,413 | 3,558 | 3,743 | 19.3% | ||
Source: Statistics of Communications Common Carriers (SOCC), Table 2-10 for 1991 to 1995
In other words, without making any changes in the current regime for recovering the interstate jurisdictional portion of the local loop costs, the per-minute charge associated with such costs will continue to decline.
Understandably, there is a concern that it might not be equitable to recover costs that are not traffic sensitive on a traffic-sensitive basis. PaISP shares this concern, but suggests that this problem should correct itself over time. The combination of technological advances and increased competition in the local loop should result in downward pressure on the cost of providing local service. The Commission should ensure that its price cap regulations do not prohibit these cost reductions from being passed on to consumers through lower access charges and, ultimately, a lower SLC.
In summary, the combination of increased interLATA calling and reductions in the cost of providing service should lead to a gradual reduction, and the ultimate elimination, of non-traffic sensitive access charges without requiring any change in the SLC. Thus, the Commission would be able to achieve its goal of reducing the potential inequities of having fixed costs collected through a usage-based charge, while avoiding the potentially devastating effects of increasing the price of accessing the telephone network.
In the event that the Commission disagrees with PaISP and determines that it is necessary to increase the SLC, PaISP supports the Commission's proposal to phase in any such increase. PaISP would emphasize that this is by no means a reasonable resolution of this issue. As discussed above, there is no compelling reason to increase the SLC at this time. However, if the Commission decides to the contrary, any change in the SLC must be phased in.
Such a phase in is particularly important because, as PaISP noted, it is very likely that the combination of new technologies and increased competition will result in lower per-line costs in the near future. The Commission should ensure that it does not increase the SLC dramatically - which could result in depressing the demand for new lines - only to have to decrease the charge shortly thereafter to reflect increased efficiencies and other downward cost pressures.
In addition, if the Commission determines that the SLC should increase, then it should continually monitor the actual cost, and the reasonableness of the cost, of providing the interstate jurisdictional portion of the local loop that is to be recovered through that charge. Simply adhering to a price cap does not serve to adequately protect consumers in an era of increasing competition and rapid technological advances and cost decreases. It should not be assumed that these costs will increase over time, or even remain constant. PaISP knows first-hand that dramatic advances in communications technology are being made daily. As competitive pressures increase, LECs will be required to implement cost-saving technologies in order to remain competitive.
Finally, PaISP strongly disagrees with the Commission's suggestion
that it might be appropriate to permit, or even require, LECs
to geographically deaverage the SLC. It must be remembered that
the SLC is just one of three methods that are used to recover
the costs of the local loop (the others being the charges for
local service and interLATA access). When all of these methods
are considered together, PaISP believe that rural customers are
already paying more per loop than urban customers. This is illustrated
in Table 2.
| InterLATA Minutes per Line | |||||||
| InterLATA Calls per Line | |||||||
| Local Calls per Line | |||||||
| End User Access Rev. per Line | |||||||
| Other Interstate Access Rev. per Line | |||||||
| State Access Rev. per Line | |||||||
| Total Access Rev. per Line | |||||||
Source: Calculated from SOCC, Tables 2-9 and 2-10 for 1995
Table 2 compares data for several LECs in Pennsylvania to a LEC that is highly urban (Bell Atlantic - District of Columbia). With the exception of Bell Atlantic-Pennsylvania, the Pennsylvania LECs tend to serve rural areas and smaller cities. As a result of many factors (probably including smaller local calling areas and different demographic patterns), it appears that customers in rural areas tend to make more interLATA calls, and have more interLATA minutes, per line than do customers in urban areas. Conversely, it also shows that customers in rural areas tend to make fewer local calls per line. Further, when the total cost recovery for access - end-user, IXC, and state - is considered, it appears that rural areas already are paying substantially more per line for access than are customers in urban areas.
The import of these facts is clear. Because of different rates for local service and a greater incidence of interLATA calling per line, LECs serving rural areas will receive more access revenue per line than LECs serving urban areas. This is consistent with the cost of the local loops which tend to be lower in urban areas than they are in rural areas.
As the Commission recognizes, section 254(e) of the Communications Act, 47 U.S.C. § 254(e), does require the Commission to make explicit any subsidies for universal service support. However, there is no clear indication that charging the same SLC in urban and rural areas will result in any subsidization of rural customers. Such a subsidy would exist only if the total cost of providing access were less than the SLC. This is unlikely to be the case under the current cost structures of the LECs. The remaining recovery of this cost is achieved through two other rates - IXC access charges and the charge for local service. As was noted above, in both instances, it appears that customers in rural and high cost areas are paying more per line toward the cost of the local loop than are customers in urban areas.
Furthermore, if as a result of paying IXC rates that include access charges, urban customers end up paying more toward access than the cost of providing that service, such a result is consistent with the intent of the Act. Section 254(g) of the Act, 47 U.S.C. § 254(g), requires IXCs to charge the same rates in rural or high-cost areas as they do in urban areas. Thus, in this instance, rather than evidencing an intent to have rates precisely recover the cost of service, Congress instead expressed an intent to encourage the nationwide availability of IXC services at comparable prices. However, as was noted above, it appears that the average access cost recovery from rural customers through IXCs may exceed that recovered from urban customers, so again there is no evidence of a cross-subsidy.
In summary, PaISP urges the Commission to retain the SLC at its current rate and to allow competition and new technologies to reduce per-line access costs. Furthermore, the growth in demand for communications service will continue to gradually reduce the per-minute charges that are levied on IXCs, serving to bring further benefits to consumers. Finally, the Commission should not permit LECs to deaverage the SLC. Such an action would further harm consumers in rural areas who already appear to be paying substantially more per line for access than are consumers in urban areas.
In these paragraphs, the Commission seeks guidance on the method for determining the SLC for ISDN lines. Of particular concern to PaISP is how the SLC will be determined for Primary Rate Interface ("PRI") ISDN lines. Through new technology that has become available during the past several months, ISPs can use PRI lines to provide high-quality, cost-effective Internet service in less populated areas. The combination of this technology and the PRI line have enabled several ISPs in Pennsylvania and elsewhere to bring the Internet to smaller communities.
One important concern, though, is that for many LECs (including all of those that serve Pennsylvania), ISDN technology is still in its infancy. This means that the costs of installing and maintaining the service are significantly higher than they will be once the technology has been more widely deployed. Stated differently, the costs for PRI installations that appear on a LEC's books reflect a great deal of start-up costs and other inefficiencies that would not exist if the technology were being fully deployed. Therefore, it would be extremely prejudicial to establish the SLC for PRI installations based on the current cost structure for PRI.
Indeed, throughout Pennsylvania, PaISP is being told that, where PRI is even available, there are often substantial waiting periods for new PRI installations. In addition, ISPs have experienced many instances where the initial installation did not work properly, requiring repeated service calls to have the PRI function properly. All of these problems are consistent with the start-up nature of this technology and result in much higher costs than would be the case for a more mature technology.
Moreover, establishing the SLC based on the comparative cost of the technology would not provide any incentive to the LECs to improve their efficiency and training on the use and installation of this technology or otherwise reduce the cost of providing this service. If the SLC revenue were tied to the level of actual costs being incurred by the LEC, then as soon as the LEC became more efficient, it would be required to reduce its charges to customers. This would remove much of the incentive that a LEC would have to become more efficient and lower its costs.
Instead, PaISP suggests that the SLC for PRI installations should be based on the physical nature of the facilities. At present, an SLC is charged for each derived channel - 23 channels in the case of PRI ISDN. These charges greatly exceed the relative cost to the LEC of providing PRI service, even under the current, inefficient level of costs (see paragraph 70 of the NPRM). Moreover, as the Commission notes in paragraph 69, placing such high fixed costs on PRI installations will depress demand for the service that, in turn, will make it that much more difficult for the technology and associated costs to mature.
Therefore, PaISP suggests that the SLC for ISDN service should be based on the physical nature of the facilities. In the case of PRI ISDN service, that would be two pairs of copper wires, resulting in the payment of two SLC charges for each PRI installation. Such a rate structure would provide a readily measurable quantity of facilities over which to assess the SLC and would not be based on inefficient and start-up cost levels or other measures that are completely unrelated to the cost of providing the service.
Moreover, from the cost data that are summarized in paragraph 70 of the NPRM, this would price the SLC for PRI ISDN at twice the SLC for Basic Rate Interface ("BRI") ISDN service. BRI is provided over one pair of copper wires, while PRI uses two pairs of wires. PaISP believes that as PRI becomes a more mature technology, its local loop costs will approach roughly twice the level of cost for BRI. With current BRI local loop costs roughly equal to the equivalent costs for an analog line, it appears that setting the SLC for PRI at twice that rate would be reasonable and would provide the LECs with the appropriate incentives to improve their efficiency and training in the installation and maintenance of this service.
In these paragraphs, the Commission seeks guidance on the advisability of removing high-capacity services from price cap regulation. In areas where there is real competition among multiple suppliers for providing such services, PaISP would have no objection to allowing the market to set prices. However, it must be emphasized that real competition among multiple suppliers does not exist today for the ISPs in PaISP.
Furthermore, it is not enough to have just one or two alternative providers in an area. As we have seen repeatedly in other areas, replacing a monopoly with an oligopoly does not guarantee that the market will work competitively. Indeed, it is very likely that the suppliers will walk in lock-step with each other and consumers will lose the protection of regulation without receiving the benefits of competition. This is the worst of both worlds and must be avoided.
There may be areas where there are multiple competitive suppliers of high-speed access services, but they are certainly not in Pennsylvania's small cities and rural areas. It is imperative, therefore, that any move toward deregulating these services must examine relevant geographic areas. PaISP would suggest that the relevant geographic area must be small enough to identify the customers who require such services. Given the fact that many ISPs, schools, hospitals, businesses, and manufacturing facilities are finding it increasingly important to obtain high-speed communications facilities, PaISP suggests that this analysis should be conducted for each central office.
PaISP understands that it may appear burdensome to require a competitiveness analysis for each central office, but that is where the customers must purchase their facilities. It is of no use to an ISP, for example, to say that competitive options are available within the county, LATA, state, or other broad geographic region. The ISP, or any other business, must do business where they are located, not in some other community or location. If a competitive option is not available at the customer's location, then there is no competition for that customer.
The Commission must understand that we are no longer in an era when only very large, multinational corporations have high-speed communications facilities. Today, thousands of small businesses (including many ISPs) rely on these facilities as an essential part of doing business. These smaller businesses predominantly have facilities that have been purchased from their LEC. If the LEC is permitted to deregulate its prices, without the presence of real and meaningful competition, then those customers with existing facilities will have no place to go and no means of protection from abusive pricing by the LEC.
In short, PaISP understands the need for the Commission to step back and allow market forces to operate. But the Commission must ensure that, before it deregulates prices, that the customer has real competitive choices available to it. This means that the analysis must be done locally and must ensure that multiple providers are able to provide the necessary services.
In this portion of the NPRM, the Commission seeks comment on proposals to charge for "terminating access" - that is, receiving a telephone call. These paragraphs of the proposal seem to be based on some economic projections that charging for terminating access will somehow encourage competition for the right to be a customer's terminating access provider. With all due respect to the Commission and the authors of these proposals, this makes no sense to PaISP.
When someone receives a telephone call, particularly a toll call, there is an expectation that the party who initiated the call is responsible for paying for it, unless some specific action has been taken by the receiver of the call (such as accepting a collect call or purchasing an 800 or 888 number). PaISP cannot believe that the Commission is seriously considering having consumers pay to receive calls from telephone solicitors, junk faxes, wrong numbers, and other unsolicited, nuisance calls.
Of course, PaISP understands that another, presumably more serious, option under consideration is simply a question of allocating the cost of access between terminating and originating providers. However, if the goal of such a proposal is to lead to competition for the right to terminate access, it would require consumers to be able to select not only their originating service provider, but also their terminating service provider. It is unclear how any consumer would make an intelligent choice among terminating service providers or how the consumer would know when that service provider has been changed (either knowingly or through "slamming").
In short, consumers do not "purchase" terminating interstate access; they purchase the right to make and receive calls. The consumer understands that by initiating a call, the consumer may be charged and, increasingly, the consumer may choose who it will pay to provide that service. However, the consumer also understands that by receiving a call, the party who initiated the call is responsible for the charge.
Changing this system would result in a dramatic change in the way in which people and businesses use their telephone service. For example, would people still have answering machines, voice mail systems, and so on, if they were required to pay for each call received? How would the consumer distinguish between a local call (that might be received for free) and an interLATA call (that would not be received for free)? How would wrong numbers and unsolicited calls be handled?
It is no secret that ISPs receive a great many more telephone calls than they originate. Most of those calls are local, but in more remote areas consumers are willing to (or forced to) make an interLATA toll call to access an ISP or other information provider (such as a computer bulletin board service). Similarly, customers who are away from home and want to dial into their ISP service will do so through an interLATA call. Is it reasonable for the ISP, or any other business, to incur a cost when its customer calls from one location but not from another? Will people still be willing to establish computer bulletin boards if they will have to pay a fee for each call that is received? (A few years ago, the FCC converted its bulletin board from a toll-free number to a toll call because of the cost. Would the Commission be able to continue this service if it again had to pay a fee for each call that it received?)
Moreover, PaISP is concerned that charging for terminating access would give LEC affiliates a strong competitive advantage over independent ISPs. Many LECs have established affiliates to begin offering Internet access service. While information is not made readily available, PaISP believes that most, if not all, of these LEC affiliates do not receive calls over local loops. Rather, most calls from customers who want to access the Internet are terminating directly in a central office. If the Commission institutes a system where end users must pay for terminating access, it would result in independent ISPs having to pay such charges (because they must use local loops), while LEC-affiliate ISPs would not have to pay such charges. This would be highly inequitable and would place small, independent ISPs at a serious, competitive disadvantage.
Simply, it is one thing to allocate costs among IXCs; it is quite another to impose new charges on the consumers of communications services. PaISP has no position on any efforts to reallocate costs and payments among IXCs, but PaISP would emphatically urge the Commission to refrain from imposing any fee on the receipt of telephone calls.
The issues raised in this section of the Commission's NPRM are particularly important to PaISP. Fundamentally, the Commission is asking: Are ISPs and other information providers more like customers of LECs or more like IXCs? PaISP submits that ISPs are consumers of communications services and are not interstate carriers.
The customers of an ISP are paying to use the ISP's computer equipment. That computer equipment stores a great deal of information that is accessed directly by the customer (electronic mail, feeds from news groups, information that other customers may store so that it can be accessed by others, information that is needed for the customer's computer to become part of the global network of computers known as the Internet.). The customers of an ISP are not buying the right to use the Internet as such; they are buying the right to use the ISP's computers, modems, and other facilities to become part of the Internet. In other words, ISPs are part of the Internet and people are paying for the right to use those facilities. The communication that takes place occurs over the telephone lines that the ISP and customer have purchased from their LECs.
Importantly, consumers are beginning to have more options for how to gain access to the ISP's facilities. While most users gain access over the public switched telephone network, more consumers are establishing direct connections to the ISP through dedicated facilities. In some locations, cable television companies are offering access to ISPs through the cable system. Providers are also beginning to offer access to ISPs through satellite-based services.
Simply, ISPs are end users; they are customers of their LEC. A useful analogy might be to consider a business such as Mailboxes Etc. That business provides customers with private "post office" boxes. It receives mail and packages for the customer from delivery services. It can also send mail or packages from the customer to a delivery service. Performing these functions does not turn Mailboxes Etc. into a delivery service; it is simply another customer of the delivery services. Customers are paying for the right to use the business's facilities; they are not paying for the right to have their mail or packages sent or received.
An ISP functions in much the same manner. An ISP has a substantial investment in facilities that enable its customers to store and transfer information. One of the ISP's major investments is in telephone lines that must be purchased from the LEC. This does not turn the ISP into a communications provider; it turns it into a large customer of the LEC.
For the 15 companies that are part of PaISP, this investment is quite large. On average, these 15 companies pay approximately $100,000 annually to their LEC. In most cases, this makes the LEC the single largest supplier with which the ISP does business. They are paying for telephone lines, ISDN facilities, high-speed data lines, and other facilities that are available only from their LEC. Indeed, unlike IXCs, ISPs purchase a large number of local telephone lines from the LEC and connect facilities directly to those local lines. The companies that are participating in these comments have anywhere from 25 to 500 local phone lines each (averaging more than 200 lines each). There should be little question that these companies are customers of the LEC.
PaISP agrees with the Commission's assessment that it makes no sense to bring ISPs into the access charge regime. As was discussed above, the SLC and access charges are designed to recover the portion of the cost of the local loop that is allocated to the interstate jurisdiction. Most calls that an ISP receives are local calls and would not be subject to interLATA access charges in any event. Recovering most of the cost of the local loop through fixed charges (the local charges and the SLC) makes sense because most of those costs are not traffic sensitive.
Generally, then, ISPs are primarily receiving local telephone calls. Such calls normally would not be subject to access charges and the Commission should not take any action that would subject such calls to access charges.
Further, as the Commission states in paragraph 287 of the NPRM, any analysis of the relative costs and benefits associated with ISPs must consider the additional costs and benefits associated with the ISP's customers. Many of those customers have installed an additional telephone line in order to access the ISP. The revenues associated with that line, as well as the frequency of use of that line, also must be considered in assessing the impact of using the public network for gaining access to the Internet. This issue and others will be explored in the Commission's Notice of Inquiry at CC Docket No. 96-263.
Also to be explored in the Notice of Inquiry will be the question of whether the costs of an ISP's local loops are substantially different from the average loop cost. PaISP believes that the average cost per loop is substantially lower when an ISP has several hundred lines running to a single location, and that location is often very close to the LEC's central office. Therefore, an analysis predicated on the average loop cost simply does not apply to customers with a large number of loops terminating at a single location. PaISP intends to explore this issue more fully in its response to the Notice of Inquiry.
At this juncture, PaISP would note that it is the responsibility of the LEC to provide service to its customers, including ISPs and those who want to gain access to the Internet. If the LECs believe that they are losing money or jeopardizing the reliability of their system by using their existing facilities for providing such service, then they must install facilities that will provide reliable service in a cost-effective manner. ISPs have made no secret of the fact that they are ISPs. Indeed, many LECs have special customer service representatives to work with ISPs to help them establish service that is reliable and cost-effective. And, of course, LECs have enjoyed large increases in revenues and profits from marketing additional telephone lines to those who want to access an ISP.
For example, Bell Atlantic reports that between September 30, 1995, and September 30, 1996, its local service revenues increased by 5.4%. Form 10-Q for Quarter ending September 30, 1996, Bell Atlantic Corp., at 10-11. That company states the reasons for this strong growth rate as follows:
Higher usage of our network facilities was the primary reason for the increase in local service revenues in both the third quarter and nine month periods. Our access lines in service grew by 3.7% from September 30, 1995. We experienced strong usage growth in both the business and residential markets. Business usage was fueled by strong growth in Centrex lines and residential growth was driven by secondary lines.
Id., at 11 (emphasis added). See also "Rebound," Communications Daily (Jan. 22, 1997), noting that Bell Atlantic, Nynex, Pacific Telesis, and SBC Communications all reported strong growth in the number of secondary residential lines added during 1996 and concomitant growth in earnings from the provision of local service.
It is now necessary for the LECs to take the actions that are required to ensure that their networks can deliver on the promises that they have made to their customers. If their networks can't handle the increased traffic, they should stop selling additional telephone lines until the networks are again reliable.
Of course, PaISP recognizes that the public network may not have been designed with Internet access in mind. But that is irrelevant. It wasn't designed with fax machines, credit card verification systems, voice mail, or numerous other communications services in mind either. There is little that an ISP can do to make the public network more reliable; that responsibility belongs to the LEC. Once the LEC has installed the equipment that is necessary, then ISPs, like any other customer of the LEC, will pay their fair share of the cost. And, if that cost is unreasonably high, then the ISPs, like any other customer, will look for lower cost alternatives.
PaISP must emphasize, though, that it is no answer for LECs to say that they want to charge ISPs more (by imposing access charges or per-minute charges on received calls) before the LECs have installed any additional facilities or taken any other actions to ensure that the network remains reliable.
This Commission should ensure that the responsibility for the reliability of the local network remains where it belongs - with the LECs. As end users, ISPs will pay their fair share of the costs, but it should not require dramatic changes in the entire structure of access fees and other charges in order to make that happen.
The Commission correctly states that it must examine the impact on ISPs if it decides to impose access charges on ISPs. In addition, the Commission should recognize that other aspects of its proposals could have a serious impact on ISPs and other small businesses that rely heavily on the communications network. In fact, several of the specific issues that are raised by PaISP in these Comments could have a very serious effect on small business ISPs, specifically:
The impact of those changes on small businesses should be examined under the Regulatory Flexibility Act and other relevant legal requirements.
In summary, Pennsylvania Internet Service Providers respectfully request the Federal Communications Commission to take the following actions:
Respectfully submitted,
Scott J. Rubin, Esq.
3 Lost Creek Drive
Selinsgrove, PA 17870
(717) 743-2233
sjrubin@ptd.net
Counsel for:
Pennsylvania Internet Service Providers
Dated: January 27, 1997
Comcation, Inc.
Doylestown, PA
CSRlink, Inc.
Williamsport, PA
Cyberia Communications, Inc.
York, PA
Dayton Computer Services, Inc. d/b/a Pathway Internet
Grove City, PA
Infobahn International, Inc.
West Mifflin, PA
LebaNet, Inc.
Cornwall, PA
Luce-McQuillin Corp. d/b/a Telerama
Pittsburgh, PA
MicroServe Information Systems, Inc.
Wilkes-Barre, PA
Net Reach, Inc.
Philadelphia, PA
Observer Publishing Co., Inc.
Washington, PA
Penncom Internet Co.
Warren, PA
SunLink, Inc.
Sunbury, PA
TradeNet, Inc.
Doylestown, PA
U.S. Online, Inc.
Mount Laurel, NJ
Westmoreland Online, Inc.
Greensburg, PA