Voice Over IP:
Policy and Regulatory Issues
By Mark Steven Katsouros
May 10, 1997
**Note the views expressed in this paper are those of the author and do not necessarily represent those of the University of Maryland at College Park
Voice Over IP:
Policy and Regulatory Issues
**Note the views expressed in this paper are those of the author and do not necessarily represent those of the University of Maryland at College Park
The purpose of this paper is to explain and discuss the use of the Internet as an alternative to traditional PSTN[1]0-based telecommunications. After a brief introduction, the paper compares and contrasts the characteristics of the Internet versus those of the PSTN in general, as well as the Internet versus Interexchange Carriers (IXCs)[2] specifically. It also examines the ACTA[3] Petition before the Federal Communications Commission (FCC), and various responses to it.
Ultimately, the paper examines universal service issues and, perhaps more importantly, access charges and reform. The paper concludes with a brief summary noting specifically that the future success of the aforementioned alternative depends greatly on technical and, more significantly, policy and regulatory factors in the future.
The Internet is an extremely complex medium that has evolved to the point where it combines elements of telecommunications, computing, broadcasting, publishing, commerce, and information services into a revolutionary, new model for global communications. Unlike the PSTN, which is circuit-switched[4], the Internet is packet-switched[5]; communication on the Internet is accomplished by sending small packets of "data" back and forth. In addition to the data itself, each packet contains an address to ensure that it is routed to the correct destination. The exact format of these packets is defined by the Internet Protocol (IP). The "data", of course, can be just about anything, including encoded, digitized "voice." Thus, "Voice Over IP" is exactly that--voice that is "packetized" as defined by IP, and sent over the Internet to facilitate telephone-like communication.
Voice Over IP, like many new technologies, has more than one name; it is also often referred to as Internet Telephony, or IT (not to be confused with Information Technology), and VON, which stands for "voice on the Net". From this point onward, this paper will use the short, non-confusing "VON" acronym in place of "Voice Over IP."
Because individual packets can take different routes across the Internet and arrive at different times, packet-based voice communication is particularly challenging. Unlike a text message (i.e., e-mail), which can eventually be reconstructed once all the packetized "pieces" have arrived, real-time voice communication requires timely delivery to avoid significant pauses, which would render it virtually unusable, or at least not nearly comparable to traditional telephone communication.
Nobody can say with any real certainty exactly when VON was "invented," but it seems that the first VON products that saw any widespread use began to appear on the market in early 1995, around the time when the World Wide Web (or just "Web") was in its early, yet exciting, stages, which brought droves of new users "on-line." VocalTec, believed to have released one of the first VON software packages ("Internet Phone") to yield "good" voice communications, and the current VON industry leader, claims that there have been over 600,000 downloads of its test software.
VON is one major example of a dominant trend affecting communications technology today: Convergence. Internet Service Providers (ISPs) are increasingly offering telecommunications applications (VON, video conferencing, and so on), while "traditional" telecommunications carriers are converting their networks to provide digital transmission and packet-switching architectures (ADSL[6], ATM[7], and so on).
While VON is a challenge in the technical sense, the trend to technological convergence has created quite a challenge in the regulatory sense; it is the policy and regulatory challenges that are the focus of this paper.
As mentioned previously, there are significant differences between the Internet and the PSTN, the primary technical difference being their switching architectures. This difference in switching architectures yields other significant technical differences. The Internet uses dynamic routing (based on non-geographic addressing) versus the PSTN, which uses static switching (based on geographic telephone numbering). Furthermore, the Internet's "intelligence" is very much decentralized, or distributed, as applications are delivered by client[8]-oriented software versus the PSTN, which bundles transport and applications, resulting in the medium's intelligence residing at central points in the network.
The Internet's decentralized architecture is extremely significant, as it promotes flexibility, resulting in the rapid development of new technologies and new forms of competition. But also significant is the paradoxical relationship between the Internet and the PSTN. Since many Internet users connect to the Internet via a modem[9], which utilizes a traditional (often supplemental) phone line, and since the majority of the Internet is carried over IXC lines, most Internet applications are likely to be seen as another source of revenue to the traditional telecommunications carriers. Yet VON specifically poses an obvious potential competitive threat, which is likely to be healthy for the traditional telecommunications industry.
Also paradoxical is the "regulatory" difference between the two. The PSTN has traditionally been regulated, though policy makers are attempting to make its markets more competitive. In contrast, the Internet's markets have traditionally been competitive, and are currently being considered to be subject to some regulatory policies, again largely as a result of the convergence of their markets. These regulatory issues are discussed in detail in subsequent sections.
Finally, there exist some obvious cultural differences as a result of the different architectures. The PSTN contains "settlements" that charge each other to exchange traffic, while the Internet is, for the most part, void of usage-sensitive charges as they simply do not apply to the Internet's "peering agreement" model.
Until recently, VON required that users at both ends of the conversation each have a computer with some type of connection to the Internet. This significantly limited, obviously, the number of people actually equipped to take advantage of this technology. However, the next generation of VON applications, called "phone gateways" (already being promoted by several companies including VocalTec, Netspeak, Analogics, and Lucent Technologies), will allow users to utilize VON technology using only regular telephones, as shown in Figure 1, without the need for specialized equipment at the end-points.
This configuration, of course, has tremendous implications, as it "hides" the underlying technology from users and, more importantly, makes it accessible to practically everyone. And yet, as long as the Internet is part of the equation, even barring newer, better, data-compression algorithms, the potential for congestion still gives traditional dedicated-circuit-based communications a decisive reliability and quality of service advantage, at least for now.
VON vendors are already working on implementing the Resource Reservation Protocol (RSVP), which carves out bandwidth across the Internet for time-sensitive traffic. RSVP was designed for multimedia over IP, but would work as well for voice. With it, routers signal each other to request a clear path through a network and set it up. The major drawback: It is possible to get so many such requests that the routers cannot honor all of them.[11] And, of course, this is much more likely with the existence of phone gateways.
Regardless of these various differences, should the IXCs be concerned about the potential competitive threat?
In March 1996, ACTA filed a petition with the FCC, which sought two principle actions:
"There is something fundamentally wrong, from our members' perspective, that somebody can talk over the Internet for free. They're giving away our product," said Charles H. Helein, general counsel of ACTA.[12] Specifically, ACTA claims that the petition is for preserving a so-called "level playing field" between traditional long-distance telephone companies and the Internet. Of course, in actuality, ACTA's true objective is to protect a regulatory anomaly allowing arbitrage--technically known as "resale"--from assault by a new, technologically advanced form of competition. Revealingly, ACTA did not ask that the FCC deregulate circuit-switched telephony in response to the advent of VON, but rather that it regulate the latter.[13]
ACTA's regulatory protectionism disguised as a "public interest" challenge does raise a number of interesting issues. Among them are questions as to how to treat VON software manufacturers, how to identify "voice" packets (a virtually impossible task since there are an infinite number of ways to encode them), and more importantly, who the "carrier" is of a VON phone call. Since there is no single long-distance carrier responsible for a user's VON phone call, but instead, many ISPs, again in a "peering arrangement", it would seem impossible to apply traditional regulatory standards to the Internet model.
"In a sense, what these guys are asking is to declare the entire software industry as a telecommunications carrier. I think that's a little far-fetched," exclaims Elon Ganor, CEO and founder of VocalTec.[14]
Netscape Communications Corporation ("Netscape"), manufacturer of one of the most popular Web browsers (Netscape Navigator), which comes complete with a "Plug-in" VON application called "CoolTalk", in its response to the ACTA petition, urged the FCC to settle the regulatory status of Internet "telecommunications" functionalities by:
Local Exchange Carriers (LECs) are mainly in agreement with the ACTA petition, in that they would obviously like for VON providers to be required to pay the same rates to LECs as do "other" switchless long-distance resellers.
IXCs vary greatly on their strategies toward the Internet, which is likely the result of the varying degrees to which they view VON as an opportunity or threat. For smaller carriers, who have more to lose because they carry little if any Internet traffic, VON is most certainly viewed as a threat. According to ACTA's Helein: "The carriers with smaller revenues will be sooner impacted by the dislocation and unfairness of the technology than someone like AT&T."[16]
Should ISPs be required to contribute to the universal service fund ("USF"), even though, unlike long-distance and other telecommunications firms, they derive no benefit from POTS subsidies? On the other hand, should traditional telecommunications carriers alone be required to subsidize Internet access for schools, libraries, and hospitals?
Section 254 of the 1996 Act extended the policy that all "telecommunications carriers" that provide "interstate telecommunications services" must contribute to the universal service fund ("USF"), which supports ubiquitous and affordable telephone service for socially favored consumers (e.g., rural and low-income subscribers). Additionally, Section 254(h) directs the FCC to "enhance access . . . to advanced information services" for schools, libraries, and hospitals, which has resulted in intense pressure from traditional telecommunications carriers, as well as the popularity of the aforementioned access enhancement, to classify the Internet as a "telecommunications service" that is subject to contribute to the USF.
However, on November 8, 1996, a federal-state joint board convened by the FCC issued universal service recommendations that recognized the risks associated with treating the Internet as "telecommunications". Instead, the joint board adopted a suggestion by Netscape to address school/library subsidies under a portion of the Act that is not telecommunications-specific. According to the board:
We recommend that the Commission adopt a rule providing that discounts for Internet access, as defined below, shall be available to schools and libraries pursuant to section 254(h)(2)(A). As explained by Netscape, Internet Service Providers (ISPs) and on-line service providers (OSPs) that also offer Internet access "rely to a large degree on existing telecommunications carriers for the underlying transport facilities that constitute the Internet's backbone, as well as for local loop connections to individual Internet servers and users." Any attempt to disaggregate the network transmission component of Internet access from the information service component could serve to undermine the competitive forces that currently characterize the Internet access market at this time.[17]
On May 7, 1997, the FCC's final decision on universal service was announced: The FCC will let phone companies increase the $3.50 maximum "subscriber line charge" (SLC) that appears on monthly phone bills to $5 beginning in January 1998 for each additional phone line in a home. The rate will then increase by $1 per month every year, plus an increase equal to the inflation rate, until it reaches a cap of $9.[18] This decision could certainly be viewed as a way for the USF to be indirectly subsidized by the Internet, since additional lines in homes are often acquired and used for Internet access.
Access charges are those fees paid by IXCs to LECs to carry the first and final legs of long-distance calls. These access charges account for as much as 60% of the costs of long-distance telephone calls, though it is common knowledge that the actual cost of access is far less[19], with much of the difference accounted for by universal service subsidies. In December 1996, the FCC launched a comprehensive proceeding to reform access charges in an effort to promote economic efficiency and local competition.[20] As mentioned previously, the FCC's final decision on universal service was just announced, so the process of "de-linking" access and universal service is well on its way.
As part of the announcement, the FCC cut $1.7 billion out of the $23 billion in annual access charges. The FCC further predicts that competition and improved productivity will cut the annual payment by about $18.5 billion by the year 2002.[21] In the meantime, however, the war between the PSTN and the Internet continues. The LECs contend that increased Internet "dial-up" usage is creating "congestion" of the PSTN (because average connection times for Internet usage are substantially longer than typical telephone calls, which average 3-5 minutes in duration), and therefore increase the LECs' switching and network engineering costs.
However, a recent study by Economics and Technology, Inc. (ETI) of the effect of Internet use on the nation's PSTN suggests that various LEC studies submitted to the FCC greatly overstate the severity of the congestion that may be caused by data traffic and ignore significant sources of revenue. For instance, in a recent study by Pacific Bell, one central office (out of 790 California central offices) was mentioned specifically as an example where data traffic constituted a significant fraction of total switch traffic, and the switch was located in Silicon Valley, which may well have the largest concentration of computer industry professionals in the world.
According to John Britton, a Pacific Bell spokesman, the media reports of an impending meltdown are greatly exaggerated:
The prospect of the network being overjammed would never happen. . . . The problem is only that one location [i.e., the Silicon Valley switch] during the evening. The state's phone network is not in trouble because of the Internet.
Raymond F. Albers, the Chairman of the Network Reliability Steering Committee of the Alliance for Telecommunications Industry Solutions (ATIS), an organization of incumbent LECs, also responded to the ISP issue in November 1996, by stating that it was unlikely that Internet usage could cause "reportable" PSTN outages.
Perhaps most important of all is the fact that since the growth in additional access lines has roughly paralleled the growth in the number of Internet users (see Figure 2), at least a portion of those additional lines (and the revenues derived from them) are attributable to subscribers who use them exclusively or primarily for calling Enhanced Service Providers (ESPs) / ISPs. The study performed by ETI indicates that the incremental revenues inuring to the LECs from these additional residential access lines for the 1990-95 period have exceeded $3.5 billion nationally.[22]
It would seem obvious that the FCC agrees with this scenario, as the recent authorization to increase SLCs for additional residential phone lines clearly reflects.
Of course, VON--especially as facilitated by phone gateways--would likely yield connection times similar to those of traditional voice telephone calls and, hence, would have an insignificant impact on the PSTN. This, for the most part, completely eliminates the congestion issue, at least as it relates to VON (versus the Internet as a whole). In fact, given that the technological advances of Internet applications seem to far out-pace those of traditional telecommunications applications, one could even argue that VON has the potential to actually lessen the burden placed on the PSTN by traditional voice traffic.
To summarize, many questions regarding the policy and regulatory issues of the Internet, and VON specifically, remain unanswered. However, the FCC itself has said that government policy approaches toward the Internet should start from two basic principles: avoid unnecessary regulation, and question the applicability of traditional rules.[24] This would seem paramount as the regulatory rules and models applied to the PSTN simply do not fit the radically different technologies, architecture, pricing systems, and culture of the Internet.
Furthermore, it seems clear that VON is likely to increase competition by reducing barriers to entry in the long distance voice market as well as decreasing transport costs. It might even be logical to assume that some IXCs may someday evolve to become strictly Internet Access Providers (IAPs), as phone gateways could drastically decrease their revenue in the long-distance voice market, while increasing their revenue in the IAP market. At the very least, it seems that IXCs have little to lose, since they are "on both sides of the fence," and what they lose from long-distance voice revenue they will seemingly gain from Internet Access.
ISPs are LECs must realize that they have a strange partnership--one in which each is the other's competitor and customer. Pricing systems that promote efficiency in terms of network development and use, are in the interest of both ISPs (who depend on the PSTN to reach their customers) and LECs (who derive significant revenue from ISPs). An increase in the cost of Internet access via the incumbent LEC networks will surely result in users looking elsewhere (competitive LECs, cable modems, wireless access, and so on).[25]
The future success of VON, as well as that of most other Internet
applications, even those not yet conceived, depends greatly on
technical factors. However, it seems clear that, as always, technology
will continue to evolve, and VON will eventually (more likely
sooner than later) be a widespread reality. More significant
(and more likely to slow down VON's full-scale deployment) are
the regulatory factors that the pace of technological change
makes so confounding to policy makers. But in spite of all the
smoke and confusion, VON is certain to emerge on a grand scale,
especially if the FCC sticks to its guns, and lets the Internet
and its applications evolve as they have in the past: Without
unnecessary regulation; without unnecessary rules.
Business Communications Review September 1996 Regulation: It Ain't Over Pg. 51
Computer Telephony March 1997 Audio, Video and Data Conferencing Over the Internet Pg. 192
CTI Vol. 2 (1) Internet Telephony: Separating Facts From Hype Pg. 76
CTI Vol. 2 (2) Internet Users Beware Pg. 8
Digital Tornado: The Internet and Telecommunications Policy Working Paper by Kevin Werbach, http://www.fcc.gov/Bureaus/OPP/working_papers/oppwp29.pdf (March 1997)
The Effect of Internet Telephony on the Long Distance Voice Market Working Paper by Andrew Sears, http://rpcp.mit.edu/~itel/ldeffect.html (Initial draft 11/95, last revised 10/4/96)
The Effect of Internet Use on The Nation's Telephone Network Lee L. Selwyn and Joseph W. Laszlo, http://www.internetaccess.org/eti_toc.htm (January 22, 1997)
Federal Communications Commission CC Docket Number 96-45 Universal Service Joint Board Recommended Decision
Federal Communications Commission CC Docket Number 96-488 Access Charge Reform (NRPM) and Internet Access & Information Service Provider (Notice of Inquiry; NOI) http://www.fcc.gov/isp.html
Federal Communications Commission Fact Sheet The FCC, Internet Service Providers and Access Charges http://www.fcc.gov/Bureaus/Common_Carrier/Factsheets/ispfact.html
Federal Communications Commission Local Competition Order
Federal Communications Commission NEWSReport No. DC 96-113ACTION IN DOCKET CASE (December 24, 1996) Commission Initiates Proceeding to Reform Interstate Access Charges http://www.fcc.gov/Bureaus/Common_Carrier/News_Releases/1996/nrcc6088.txt
Federal Communications Commission Notice of Proposed Rulemaking, Third Report and Order, and Notice of Inquiry http://www.fcc.gov/Bureaus/Common_Carrier/Notices/fcc96488.txt (Adopted: December 23, 1996, Released: December 24, 1996)
Federal Communications Commission Web Site http://www.fcc.gov
Internet Access Coalition FCC Notice of Proposed Rule Making (NPRM) http://www.internetaccess.org/fcc.htm
Internet Access Coalition Home Page http://www.internetaccess.org/
Internet Economics Book by Joe Bailey and Lee McKnight
The Internet Telephony Consortium Internet Telephony Watch http://itel.mit.edu/acta/acta.html
The Internet Telephony Consortium Web Site http://itel.mit.edu/
An Introduction to Internet Economics Lee W. McKnight and Joseph P. Bailey, http://www.press.umich.edu:80/jep/works/McKniIntro.html (March 1995)
Joint Opposition of Netscape Communications Corporation, Voxware, Inc. and Insoft Inc. RM No. 8775 (May 8, 1996)
MICOM Communications Corporation V/IP Frequently Asked Questions V/IP Adds MICOM's Voice to Your IP Network http://www.micom.com/press/VIPintro.html
Netscape Communications Corporation News & Reference Netscape Opposition to ACTA Petition on Internet Telephony http://www.technologylaw.com/techlaw/acta_comm.html
Network World October 28, 1996, Pg. 47
Network World November 25, 1996 Voice over IP is sounding better Pg. 24
PRNewswire March 4, 1996 FCC Petitioned to Stop Misuse of the Internet! http://www.pulver.com/von/vonyes/actapet.htm
Regulation of the Communications Functions of the Internet: "A War Between Two Worlds" Presentation by Glenn B. Manishin, http://www.technologylaw.com/techlaw/iworld/iworld1.html (December 13, 1996)
Telephone Conversation between FCC Chairman Reed Hundt and Delta Three CEO, Mr. Elie Wurtman http://www.virtual-voice.com/About/d3trans.html (March 14, 1997)
The Virtual Voice What's New at The Virtual Voice? http://www.virtual-voice.com/About/whats_new.html
Wall Street Journal March 3, 1997 Motorola to License VocalTec Software
The Washington Post March 8, 1996 Freebie Heebie-Jeebies Pg. F1
The Washington Post May 8, 1997 FCC Approves Restructuring Of Nation's Telephone Rates Pg. E1
The Washington Post May 8, 1997 $2 Billion in Discounts
to Aid School's Access to Internet Pg. E1