Technical and Economic Assessment of Internet Protocol Version 6 (IPv6)
Stakeholders agree that the market should and will be the primary driver for the adoption of IPv6. Some are concerned, however, about the implications for U.S. competitiveness if America lags behind other nations in the adoption of IPv6. Actions by governments in Asia and Europe to promote deployment of IPv6 in their countries suggests that those governments believe that there may be “first mover” advantages from early adoption of the new protocol. There have been reports that some foreign governments see an opportunity to leverage the development and deployment of IPv6 to strengthen their position in information technology and, specifically, in Internet equipment, software, and services.[1] U.S. stakeholders worry that if the United States loses its current technical and market leadership in the Internet sector, recapturing that position will be difficult.[2]
Companies that first adopt a particular technology (“first movers”) may in some circumstances have the ability to create barriers to subsequent entry or to influence the adoption decisions of other companies. By so doing, the first mover may be able to dominate the markets associated with that technology and generate monopoly profits.[3]
First mover advantages arise from three general factors: (1) technology leadership, (2) preemption of scarce resources or assets, and (3) an ability to “lock in” users due to the high costs of switching to alternative technologies or products.[4] Because the principal resources underlying the Internet are information and human capital, and because markets for those resources are international, open, and highly mobile, it is unlikely that the second factor will confer any first mover advantages with respect to IPv6. On the other hand, first movers may gain learning and experience that will enable them to exert technology in the IPv6 market. If all companies are required to move along similar learning curves in developing and deploying IPv6 products and services, early movers may be able to sustain knowledge advantages, leading to lower costs or higher quality products and services.
Knowledge and experiences gained by early adoption could also allow first movers a competitive advantage in providing implementation services (potentially tied to hardware and software purchases). As in the early years of IPv4, many unforeseen issues are expected to arise during real-world implementation of IPv6. Resolving these issues will be an important factor in customers’ purchase decisions. Hardware and software suppliers that develop this expertise early will be in a better position to attract and retain customers.
First-movers in IPv6 markets could also create high switching costs that “lock in” users (i.e., that prevent or deter users from abandoning the first-mover’s product in favor of a subsequently-offered substitute) and thus raise barriers to entry for latecomers to a market. Once users have invested resources in training staff to maintain and trouble shoot particular hardware and software systems it may become costly to switch. Familiarity and risk associated with change tend to support first-mover advantages.
Some industry experts question whether first movers will be able to capture sustainable competitive advantages in Internet markets. In the applications markets, for example, the rapid pace of technology advances makes sustaining first-mover technology or information advantages difficult.[5] In addition, the short life expectancy of Internet technologies and regular replacement of hardware and software applications reduce lock-in costs.
Information spillovers also work against first-mover advantages.[6] Deployment costs are typically more costly for innovators and early adopters of new technologies compared to the costs for imitators and later adopters. If U.S. companies are able to learn from the international community’s early IPv6 adoption activities, this may lower the U.S. deployment costs and lead to competitive products and services with lower entry costs. Empirical research has shown that failure is also a common outcome for first-to-market participants.[7]
Judging from the published literature and the RFC comments, U.S. stakeholders are aware of first-mover concerns, but disagree about the potential competitive effects of the United States lagging behind other nations in deployment of IPv6. At this time, most markets for IPv6 products and services are in their infancy. Until applications and services markets begin to mature, it is not possible to determine whether efficiency gains or learning curve effects will generate sustainable first mover advantages. The following information would be needed to investigate fully potential first mover issues:
· the transferability of early lessons learned from one company or system to another;
· efficiency gains from economies of scope in applications markets (e.g., bundling and product tie-ins); and
· user lock-in costs and the availability of interoperability solutions to minimize these costs.
An important point for this analysis is the fact that first-mover strategies are usually discussed with respect to the benefits and costs of innovation in applications. However, the issue here is the evolution of a critical infrastructure -- a standard. Standards provide several functions that enable innovation: (1) reducing variety (i.e., one standard versus several incompatible protocols) and thereby presenting larger potential markets; (2) providing information (e.g., format and timing of message transmissions) and thereby reducing the costs of innovation; (3) assuring quality (e.g., accuracy and assurance of message delivery); and (4) assuring compatibility/interoperability (e.g., seamless integration of sub networks and applications) and thereby realizing network externalities.[8]
This section discusses transition issues and their potential impact on competitiveness for U.S. vendors of IPv6 hardware and software applications and for U.S. users that rely on the Internet to provide their products and services.
Several commenters voiced strong concern that other countries are advancing IPv6 at a much faster rate than the United States, and that without government action to stimulate or assist U.S. deployment, the United States could lose its leadership role in Internet issues.[9] Another commenter indicated that a lack of U.S. technical experience in new IPv6-based equipment and applications development could put domestic firms at a disadvantage, as other countries would be able to work without NATs and other IPv4 work-arounds.[10] Other commenters focused on resource constraints; if transition to IPv6 in the United States lags behind the international community, U.S. vendors will need to allocate resources to support both IPv4 and IPv6.[11] As a result, U.S. firms would have fewer resources to devote to IPv6-only products and services.
The requirement by potential first movers that the aforementioned economic levers from standardization be in place can act as a barrier to innovation because the profit potential is significantly reduced. Conversely, the cost of implementing a new standards infrastructure (as discussed in Section 2.2 supra) is substantial and not returnable to individual private firms. This situation raises a “chicken-or-egg” problem. ITEF has attempted to eliminate this concern by creating mechanisms to promote interoperability between IPv4 and IPv6 networks, thereby facilitating a gradual transition strategy.[12] However, the complexity of IPv6 implies the possibility that sequential and isolated implementations will not provide sufficient benefits of the types listed above to provide incentives to potential first movers. Thus, the possibility must be considered that a slow migration will also mean a slow realization of the potential greater benefits from IPv6. More concerted efforts in competitor nations to more rapidly implement IPv6 could provide advantages to those nations’ domestic firms.
However, Internet equipment and software manufacturers already compete internationally and therefore should not be significantly affected by lagging U.S. adoption.[13] U.S. vendors will continue to develop and sell IPv6 products and services in the global marketplace even in the absence of U.S. adoption. Most of the major U.S. suppliers of hardware and software applications serve international markets and have subsidiaries in Asia and Europe. Technology developed and knowledge gained through subsidiaries’ participation in regional markets where IPv6 activity is high should be transferable to the United States when U.S. deployment increases. This scenario will only hold, however, if learning economies in domestic markets are not essential to U.S. firms becoming competitive globally.[14]
An important issue related to long-run U.S. competitiveness is the shift of intellectual (human) capital from domestic locations to foreign subsidiaries of U.S. companies. If much of the IPv6 research is conducted in Asia and Europe, this could lead to a trend of high-tech Internet applications development capabilities (and jobs) migrating to these regions. Even if the activities are associated with U.S. firms, this form of “outsourcing” could lead to shifts in intellectual capital and have implications for where the next generation of Internet capabilities is developed.
U.S. companies that rely on the Internet to provide their products and services should not be substantially affected if the United States lags other parts of the world in IPv6 deployment. Large portions of the Internet will continue to operate in IPv4 and transition strategies have been, and will continue to be, developed to ensure interoperability between IPv4 and IPv6.
Several commenters suggested that government incentives could be used (e.g., tax breaks or grants) to help offset transition costs.[15] In their view, those incentives could also be beneficial in the early stages of U.S. deployment to stimulate IPv6 adoption in enterprises.[16] However, other stakeholders have warned that government incentives would be unwise because they might skew the path of technology development or interfere with ongoing activities in the commercial marketplace.[17] These stakeholders prefer that government simply participate in the market by adopting IPv6 when it is beneficial to its needs.
Currently no productivity benefits for corporate or industrial uses are associated with operating IPv6 versus IPv4, and higher costs may be associated with early adoption of IPv6.[18] When more advanced IPv6 applications become available that represent efficiency gains, U.S. companies will be well positioned to take advantage of these opportunities. As discussed above, no market failures seem to exist that would limit rapid deployment of IPv6 once future applications emerge.
Many of the issues related to general interoperability are even more relevant when discussed in an international context. Specific examples of international interoperability issues include different levels of conformance and implementation strategies across regions and legal and privacy implications of encryption restrictions across countries.
International interoperability issues associated with dual IPv4 and IPv6 network capabilities should be minimal because IPv4 is well established globally and can be used as a network foundation; however, interoperability between IPv6 applications needs to be tested more extensively in an international context.
Of particular significance to an international discussion is the impact of interoperability, or a lack thereof, on U.S. competitiveness both in Internet hardware and software and in other industries. The following sections address these issues.
International interoperability generated by standardization tactics of individual countries can create market barriers for U.S. hardware and software suppliers by raising the cost for U.S. companies to compete in international markets. One such example is the current development of China’s wireless standards. Until mid-May 2004, China intended to implement a new encryption standard for wireless communications and announced that verification of Wireless LAN Authentication and Privacy Infrastructure (WAPI) compliance would be part of its compulsory registration process for electronics.[19]
WAPI was portrayed as China's solution to the problem of securing wireless communication. However, multinational companies suggested that WAPI had security holes and gaps that would have created a burden for manufacturers who would have needed to meet one standard for China and another for the rest of the world.[20] The WAPI development is just one example of how a country could use a standard to create a trade barrier.
Even in a world where the international community cooperates to minimize problems, parallel ongoing development activities in Asia, Europe, and America will inevitably lead to interoperability issues, and companies that are active early in the process will have the opportunity to influence solutions and gain valuable experience. For example, to compete effectively in international router markets, U.S. suppliers will need to provide leading-edge support for IPv6. However, it may be more difficult to develop the needed capabilities if U.S. networks and services remain predominately IPv4-based. One commenter suggested that to compete in a global market with interoperability issues, IPv6 deployment should be encouraged domestically. As a result, American router vendors could move up the leaning curve more quickly and be competitive in international markets where IPv6 will be even more heavily (or more obviously) emphasized.[21] In other words, the usefulness of standards as a means of reducing interoperability problems, coupled with potential learning economies (a timing issue), are possible rationales for a more rapid transition to IPv6.
U.S. suppliers of non-Internet-related hardware and software should not be put at a competitive disadvantage based on international interoperability issues, according to experts. In fact, for U.S. vendors, costs would also be lower in a scenario where the early deployment issues are encountered and resolved in foreign countries.[22]
In general, an embedded base of IPv4 equipment should not preclude the United States from the benefits of foreign IPv6 deployment, as long as there is a means to connect embedded IPv4 networks and equipment to newly-deployed IPv6 equipment when legacy application support is required. The developers of IPv6 have attempted to accomplish that goal by making IPv6 backward compatible with IPv4 via interoperability mechanisms.
However, a few commenters indicated that an embedded base of IPv4 equipment and applications could function as a barrier that would isolate the United States from the benefits of foreign IPv6 deployments and/or test beds. Forward-thinking entrepreneurs might not be able to develop new services based on IPv6 or may simply participate in the new economies emerging in other IPv6 geographies.
With respect to domestic innovation incentives, small and medium U.S. businesses have limited resources. Thus, if they encounter high costs due to partial IPv6 deployment domestically, or if foreign competition benefiting from learning economies elsewhere in the world penetrates the U.S. market, barriers to domestic innovation efforts could be significant. Incomplete deployment also may send inaccurate market signals and result in premature introduction of IPv6 products, which could be damaging to small and medium firms.[23]
Finally, in the transition to IPv6, one of the most important interoperability objectives is to ensure the security and stability of IP networks around the world. Therefore, any transition to IPv6 should move forward in a cautious and technology-sensitive way to minimize adverse effects for users. International standards development and coordination bodies should be used to vet technical issues pertaining to IPv6 migration and the coordination of interoperability issues.
[1]See, e.g., Ikeda and Yamada, note 36 supra, at 2, 12; Hain Comments at 1; Motorola Comments, at 5; Dillon Comments at 1. See also Cisco Comments at 22 (Chinese carriers may feel political pressure to showcase China as a technology leader).
[2]See Alcatel Comments at 2; Cisco Comments at 24; Hain Comments at 8; NAv6TF Comments at 6-7.
[3]See P. Stoneman, The Economics of Technological Diffusion (2002).
[4]See M. Liberman and D. Montgomery, “First-Mover Advantages,” 9 Strategic Mgmt. J. 41 (1988).
[5]See David Needle, “The Myth of the First Mover Advantage,” siliconvalley internet.com (Apr. 5, 2000), http://siliconvalley.internet.com/news/article.php/3541_333311.
[6]The term “spillover” refers to the fact that some benefits of a particular economic activity (e.g., R&D) frequently accrue (“spill over”) to parties other than the one that originally undertook the activity. “Information” or “knowledge spillovers” result from the movement of information from the originating firm to other producers (e.g., through publication of the originating firm’s basic research, through “reverse engineering” of the originating firm’s product by other firms, or by the movement of employees from the originating firm to other organizations). “Market spillovers” result when the operation of the market for a new product or process causes some of the benefits thereby created to flow to producers and consumers other than the innovating firm. See, e.g., Bronwyn Hall, "The Private and Social Returns to Research and Development", in B. Smith and C. Barfield, eds., Technology, R&D, and the Economy 140 (1996); Adam Jaffe, "The Importance of ‘Spillovers’ in the Policy Mission of the Advanced Technology Program," J. Tech. Transfer 11 (1998); Zvi Griliches, "The Search for R&D Spillovers", NBER Working Paper No. 3768 (Nat’l Bur. Economic Res. 1991).
[7]See G. Tellis and P. Golder, “First to Market, First to Fail? Real Causes of Enduring Market Leadership,” 37 MITSloan Mgmt. Rev. 65 (1996).
[8]See Gregory Tassey, “Standardization in Technology-Based Markets,” 29 Res. Pol. 587 (2000).
[9]See Alcatel Comments at 4; Hain Comments at 17-18; Lockheed Comments at 5; NAv6TF Comments at 6-7.
[10]Interview with representatives of Internet2 (designing the next generation of Internet applications will be simpler in IPv6 because developers will not need to build on the more than 20 years of work-arounds embedded in IPv4).
[11]See Alcatel Comments at 4 (R&D activities could be diluted because new products and services will need to be dual protocol compatible, potentially causing U.S. companies to lag behind in developing next generation IPv6 applications).
[12]See, e.g., Hain Comments at 10-12.
[13]See Cisco Comments at 11.
[14]This last point has some historical basis. Beginning in the 1980s, Japanese electronics firms regularly introduced new products into the Japanese domestic market first in order to gauge consumer reaction and gain production experience (lower cost) before committing to head-to-head competition in global markets. See Gregory Tassey, The Economics of R&D Policy (1997).
[15]See Motorola Comments at 2; NAv6TF Comments at 46.
[16]See Cisco Comments at 16; Motorola Comments at 2; NAv6TF Comments at 44.
[17]See Microsoft Comments at 12.
[18]Interview with John Streck, Centaur Labs (May 2004).
[19]See Grant Gross, “China agrees to drop WAPI standard,” NetworkWorldFusion (Apr. 21, 2004), http://www.nwfusion.com/news/2004/0421chinaagree.html.
[20]See “China Promotes New Wireless Encryption Standard,” PulseOnline (Dec. 2003), http://pulse.tiaonline.org/article.cfm?id=1911.
[21]See Alcatel Comments at 2.
[22]See BellSouth Comments at 6.
[23]See Cisco Comments at 16.