Communications & Convergence Review 2011, Vol. 3, No. 1, 90-100

1)

Network neutrality: Cases and perspectives from Korea Sung-Hwan Kim* Ajou University

Abstract This paper examines recent cases and discussions of network neutrality in Korea. Network neutrality has become an urgent current issue with the emergence of such applications as P2P-based CDN and m-VoIP that telcos hope to restrict or charge. We discuss the issues and perspectives generated from the cases. Questioning the dynamic efficiency of network neutrality standard, we argue that a policy beyond network neutrality is needed to let the users achieve balanced consumption of differentiated IP services. Keywords: network neutrality, internet, P2P-based CDN, m-VoIP, Korea

Introduction The year 2010 marked a crucial turning point for Korean telecommunications industry, when smartphones became quickly popular causing fundamental changes in the industry. And it was also a big turning point for the network neutrality discussion in Korea. With drastic changes of the market environment, network neutrality has suddenly turned from an important but long-term issue to an urgent current issue. The most important driver was the emergence of such new applications as P2P-based CDN (Content Delivery Network), m-VoIP, and smart TV, which threaten telcos’ revenues seriously. Telcos started to mention restricting those applications or charging additional fees to them. 1) Although there have been some works on network neutrality in Korea so far, most of them focused on finding implications of the policies developing in other countries and the Korean issue of their interests was usually concerned with opening premium IP networks to potential non-telco IPTV service providers,

which was a significant network neutrality issue of Korea back in 2007-2008. In this paper, I aim to examine the more recent cases and related discussions that have made network neutrality, for the first time, a central issue of telecom industry in Korea. I hope that views and approaches discussed in this paper, being somewhat different than those in U.S. or Europe, can provide useful lessons for the policymakers and researchers in other countries. I do not merely introduce the cases and discussions but explain them with my own perspective. Of course, I cannot appropriately represent all of the various perspectives within Korea. But, engaging in the issue of network neutrality for about 5 years, I have interacted with and been affected by various scholars, government officials, industry experts, and other commentators. So even if my perspective might not be called “the” Korean perspective, it would definitely reveal some of the perspectives that currently exist in Korea. And I have to admit that my perspective is somewhat constrained to economic one since I am an economist.

* Assistant Professor, Division of Social Sciences, Ajou University, Korea. E-mail: [email protected] 1) Kim, Lee, & Kim (2008) and Lee & Kim (2008) are the examples.

S. H. Kim / Communications & Convergence Review 3(1)

In the last few pages of the paper, I provide my own argument more clearly. Questioning the dynamic efficiency of network neutrality standard, I argue that a policy beyond network neutrality is needed to let the users achieve balanced consumption of differentiated IP services. The paper proceeds as follows. Section 2 and 3 brief about internet access service and network neutrality policy in Korea. Section 4 examines the cases of P2P-based CDN and m-VoIP. Section 5 concludes by suggesting an idea of policy beyond network neutrality.

Internet access service in Korea Broadband access service in Korea is provided by 3 national operators, KT, SK Broadband, LG U+, and many other local operators, most of which are cable system operators. With those players in action to

91

obtain and maintain subscribers, competition tends to be strong. Considering the growing competition in the market, Korea Communications Commission (KCC) announced in December 2009 that there was no SMP (significant market power) operator in the market, which means ex ante price regulation on KT’s broadband access service being lifted for the first time since 2005. And it is worth noting that the competition in the market is recently focusing on selling TPS or QPS including IP/Cable TV, VoIP, mobile services. Total number of broadband subscribers reached 16.8 million as of June 2010. And more than 80% of them are shared by 3 national operators, KT (42.8%), SK Broadband 2) (22.6%), and LG U+ (15.7%). Mobile internet service market is growing very rapidly in Korea since smartphones gained quick popularity in 2010. While the number of smartphone 3) users was less than 1 million in 2009, a news report says that, it reached 7 million at the end of 2010 and passed the 10 million mark recently.

Figure 1. Average broadband monthly price per advertised Mbit/s, USD PPP

Source: OECD, Broadband Portal (October, 2009) 2) The subscriber numbers and market shares are from KISDI Telecommunications Policy Research Division (2010). 3) SBS News (2011. 3. 25).

92

S. H. Kim / Communications & Convergence Review 3(1)

As in any other countries, internet traffic is increasing rapidly in Korea. P2P is thought to contribute to it the most. KT once reported that only 5% of the users cause 50% of the entire traffic on its network, 88% of which is of P2P file sharing (Y. Kim, 2010). Measured by average broadband monthly price per advertised Mbit/s as of October 2009 (Figure 1), broadband price in Korea is the lowest among OECD countries at USD 1.76. One important and unique constraint to providing broadband access service in Korea is the flat rate pricing. Unlike many other countries where bit/data caps are introduced with usage-based pricing for additional amount of use, Korean telcos have maintained flat rate pricing intact and will have little choice but to keep doing so for foreseeable future. A very competitive market environment is definitely one that forces it. But, it is also due to the bad experience KT underwent when it tried to introduce usage-based pricing back in 2004 and 2005. There was an extremely harsh resistance of broadband users against KT’s proposal so that “usage-based internet pricing” became literally a taboo word in Korea. Even after over 5 years passed, telcos as well as KCC are still allergic to even mentioning the word. Such hatred for usage-based internet pricing is probably due to the fact that the proportion and influence of heavy users are especially high in Korea relative to other countries. For instance, a great number of on-line gamers and middle/high school students watching on-line lectures were benefiting from flat rate pricing and responded very emotionally to KT’s proposal. It is indeed a very strong constraint in considering the solutions to network neutrality problems in Korea. Especially from the perspective of economists, the easiest and the most essential way to resolve the conflicts regarding internet traffic may be to deal with pricing. As long as usage-based internet pricing remains taboo, however, we have to do without such essential key. Anyhow, it is the unique reality in Korea we need to keep in mind in the following discussion.

Network neutrality policy in Korea So far KCC has not announced any official position for network neutrality policy. In fact, it was not even among the major telecom policy issues in Korea until recently. It may be attributed to the existence of vigorous competition in broadband access market and enough volume of broadband capacity relative to other countries. Moreover, there is some recognition that it may be unnecessary to introduce new policy or law for net neutrality because KCC already has enough power to regulate broadband network operators based on some related provisions in the Telecommunications Business Act. It does not mean, however, that network neutrality was regarded only as a trivial issue in Korea. KCC has been extremely cautious trying to avoid hasty commitments whenever asked to express its opinion towards network neutrality. They were so probably because they knew how immense effects this issue would potentially have on the future of Korea’s communications industry in the long term, if not in the near term. So while saving words, the regulator have always shown great interests in learning about what are discussed on net neutrality by the economists, lawyers, and other experts in the industry as well as the ongoing development of net neutrality policy in other countries. In some sense, telecommunications policy in Korea is already constrained by network neutrality principles even though none of those principles are clearly defined or officially approved. The policymakers and regulators these days are often concerned about potential network neutrality problems when addressing many different policy issues of telecom markets. As for telcos, although they are naturally at odds with net neutrality idea itself and take every chance to criticize it, they do not deny its growing importance and potential impact on their businesses. So there is little disagreement in Korea to the expectation that network neutrality will become one of the most significant policy issues in the future era of all-IP network while the other conventional policy of regulations for such as entry and pricing are gradually

S. H. Kim / Communications & Convergence Review 3(1)

abolished. And in 2010, as I explained in Section 1, network neutrality had suddenly turned from an important but long-term issue to an urgent current issue. It led to the operation of Network Neutrality Forum hosted by KISDI (Korea Information Society Development Institute) and supported by KCC. To my knowledge, it was the first case in Korea of officially publicizing the network neutrality issue. And it was seen by many as a preliminary stage of collecting various opinions before KCC moving to formulate a certain network neutrality policy at last in 2011 or later. The forum, chaired by professor Chon Pyo Lee of Seoul National University, consisted of the members of various backgrounds and interests, including the professors of law, economics, engineering, and the executives of telcos and on-line service providers as well as the experts from KISDI and KCC. The discussion made in the forum may have fallen short of the expectations of some observers since it did not come up with any concrete solutions to the pending conflicts. Still it was meaningful in that the experts of different views and interests gathered and taught one another to see what is problem and how we can approach to solve it.

93

Reflecting all the opinions and ideas expressed in the forum, Dr. Hee-Su Kim of KISDI proposed a draft of guidelines on network neutrality in November 2010 (H. Kim, 2010). Although it is just a draft without any legal binding, it could be a starting point for defining and establishing network neutrality policy in Korea. KISDI guidelines consist of three parts - (1) basic rights of internet users (2) internet traffic management (3) managed internet service and QoS of best-effort internet. The first part is quite similar to FCC’s 4 principles except that KISDI adds a condition that the rights of internet users do not apply to the case of violating contracts with ISPs that restrict provision for a third party or for commercial use. So I translate and summarize only the other two parts below. The guidelines state that some of the technological or economic management of internet traffic should be allowed. And the allowable types of internet traffic management are positively listed (Table 1), based on which the firms in the internet industry are asked to jointly establish more concrete and objective traffic management guidelines of their own.

Table 1. Allowable types of internet traffic management To restrict behaviors that harm the network such as DoS attacks, viruses, and spams To respond to an emergency of temporary network overload To restrict traffic sent and received by a user according to the request of that user To restrict behaviors sending illegally the files protected by copyright To restrict providing commercial-purpose telecom service via P2P file sharing without legal arrangements When received legitimate requests from national authorities to respond to public security problems, national emergencies, etc. or when necessary to enforce other laws And in other cases where it is obviously necessary to manage internet traffic

In addition to the list, the guidelines suggest 4 principles of internet traffic management as follows, • Transparency: ISPs should inform the users of their internet traffic management rules specifying the criteria, targets, and methods of traffic management

• Relevance and Proportionality: Internet traffic management should be highly relevant to resolving the noted problem and be proportional to the extent of the problem • No discrimination: Traffics of the same technolo-

94

S. H. Kim / Communications & Convergence Review 3(1)

gical properties should be treated equally. • Economic management favored: Economic management of traffic should be applied in advance of considering technological management. As for traffic management in wireless internet, the guidelines state that it should be acknowledged that the constraint of spectrum imposes limit to expanding capacity in response to traffic congestion. According to the guidelines, ISPs may provide managed internet services that ensure QoS for individual users and firms, CPs, and other ISPs and may charge additional fees or network usage costs. Such allowance of managed internet service comes with several responsibilities for QoS of best-effort internet. First, they need to make sure that providing managed internet services does not result in degrading the quality of best-effort network, under the supervision of KCC. Second, they need to publicly announce the objective information on change in quality of best-effort network they own, on regular basis. Third, ISPs need to cooperate in developing the objective index concerned with maintaining and improving the quality of best-effort internet, which should also be supported by the government.

Two cases in debate Although we can clearly see in the KISDI guidelines the efforts of adapting network neutrality ideas to the reality of Korean markets, they still look quite analogous to the policy statements of other countries in America and Europe. But, it is a starting point and network neutrality policy of Korea will develop to a far more concrete form before long. And I expect that it will be affected significantly by how we address the two recent cases that I elaborate on in this section. They raise many issues that are quite unwieldy yet still need to be resolved.

4) Inews24 (2010. 10. 22).

P2P-based CDN CDN (Content Delivery Network) is used to reduce service interruptions and save bandwidth cost in providing bandwidth-intensive and time-sensitive online services such as internet video distribution. For CDN service, edge servers are installed at strategic locations close to clients to reduce the load on the origin server (Ha, Wildman, & Bauer, 2010). To be a little different than CDN, P2P-based CDN utilizes the resources of private broadband users with P2P technology. The service providers lease the broadband lines that were originally contracted for home or office use and attach some devices to them to, in some sense, resale them to the customers of CDN service. This way they can save the server cost. As this special kind of service became popular in Korea, telcos started to react sensitively to it in 2010 and condemned the service as an unfair free ride. According to the telcos, such service causes excessive traffic without enough contribution to the network. Moreover, it substitutes for server-based CDNs, reducing the IDC (internet data center) sales of telcos by as much as 40%. For instance, they could provide the P2P CDN service using 20 broadband lines that cost only KRW 680,000 per month, which would have cost KRW 18,000,000 per month if they had chosen to pay for servers to get 2 GB backbone service. According to KTOA (Korea Telecommunications Operators Association), telcos expect to lose KRW 4) 200 billion each year for this. We could of course downplay the problem supposing that it is simply whining of telcos that are losing their old territories in due course of technology advances. However, it would still be worth considering the issue of free ride. Even if P2P CDN is of a new technology, its economy might come primarily from handing over the network cost to the telcos, rather than from its own merit or innovation. In such case, we need to be careful in applying network neutrality principles as it may not be the kind of innovation that network neutrality intends to promote. So, is it really an act of free ride without any

S. H. Kim / Communications & Convergence Review 3(1)

meaningful innovation? From the standpoint of the service providers, it is an innovative technology that utilizes cheaper unused resources. And it may be rather one of the typical arbitrage behaviors that are seen as legitimate and even necessary for market efficiency. However, telcos would argue that it cannot be justified that easily because they are unduly taking advantage of the tariffs designed originally for home or office use. Who is right? To put aside legal issues and focus on economic ones, the answer depends on whether the tariffs for private broadband access can compensate justly for the network cost incurred by such heavy uploading of P2P CDN service. If the current tariffs are appropriate and sustainable even with P2P CDN service, the service can be commended as innovative for utilizing underused resources. Given the present tariff structures in Korea, however, it is indeed hard to justify P2P CDN service in terms of economic efficiency. The broadband tariffs were designed from the beginning as simple flat rate, and remains so at least as yet, to reflect the pattern and amount of average home or office usage. And they are far from being sophisticated, not restricting excessive use nor treating upload traffic differently. Therefore, P2P CDN service may be merely exploiting the vulnerability of the current simple tariff structure rather than utilizing cheaper resources. As explained in section 2, the adjustment of tariff structure is not among the usually available options for telcos in Korea. Still, there is one possible case left for P2P CDN. Regardless of the problem of tariff structure, there may exist enough redundant capacity of broadband access network in Korea so that P2P CDN services incur in effect little additional network cost at least for foreseeable future. If it is true, it would be hard to refute its efficiency case. Then, the loss of telcos in IDC business should be endured however hurting it is. It is what always happens in market economy to lose sales to the arrival of any new competitor or substitute. Of course, to assess this hypothesis we would need more concrete and objective information on the status of broadband network capacity and

95

usage in Korea, which has not been provided yet by the telcos themselves. So what would be the solution? If the telcos can concretely show only that serious level of traffic congestion is being caused by P2P CDN, I think they could restrict the service and it can be justified as reasonable network management. Even the network neutrality policy would allow such prescription. Or if there is clearly some additional cost to network but not of such level that would justify network management, then the telcos have little choice but to modify their longstanding tariff structure by introducing the elements of usage-based pricing. However, in case the last hypothesis of enough redundant capacity is shown to be true, a resolute network neutrality policy is needed to intervene for any unfair attempts of the telcos to disturb the businesses of P2P CDN service providers. Even any attempt of adjusting tariffs should be carefully monitored so that it does not have an effect of excluding P2P CDN service providers.

M-VoIP M-VoIP (Mobile Voice over Internet Protocol) service had not been an option for mobile users in Korea until 2010 when the smartphones gained quick popularity. In July 2010, SK Telecom, the leading mobile operator in Korea, began to allow customers of All-in-One plans over KRW 55,000 to use m-VoIP services on 3G network. KT, the second largest mobile operator, followed suit in December adopting a very similar policy. So, m-VoIP is now available in Korea, but only for the users who pay more expensive flat rates. Such policy of the two telcos towards m-VoIP seems designed carefully to minimize the impact on their voice revenues that are known to account for 80% of mobile service revenues, while promoting the new flat-rate plans and responding to growing user interests in m-VoIP applications at the same time. In other words, its actual effect of benefiting the users seems little as high-value users, who are currently allowed to use m-VoIP, are among the least likely to tolerate the

96

S. H. Kim / Communications & Convergence Review 3(1)

inconsistent quality of m-VoIP (HSBC, 2010). Naturally, it brings about debates on whether KCC should intervene so that the telcos’ policy of allowing m-VoIP can be extended to all mobile users. If KCC ever decides to intervene in the matter, it may want to tackle the issue of ‘undue discrimination of users,’ an act of which is subject to ex post regulation according to the Telecommunications Business Act. Whether the current practice of the telcos unreasonably discriminates users is a contentious legal issue that may require serious discussion by law professionals. But, to the extent that the regulatory decisions on undue discrimination of users have often relied on discretionary authority of KCC so far, I suppose the decision could reflect general aspect of the issue as well, that is, how much it will benefit mobile users in general if all the mobile users are allowed to use m-VoIP or how unfair the mobile users find the current practice. The potential benefit of m-VoIP saving user cost seems obvious especially when it comes to the users who are on the plans that allow less free voice minutes with lower subscription fees. Then the remaining issue is whether the current practice is unfair and unreasonable. For users who are now quite accustomed to downloading and installing various apps on their smartphones, it is hard to acknowledge the telcos blocking one of such apps, m-VoIP. In fact, it is even more so when recognizing that many of the other apps incur more traffic than m-VoIP but are not restricted in such way. And m-VoIP service providers argue that the service is not even a free ride on telcos’ network since the users pay for data usage. Telcos of course have good reasons of their own to disagree. In their eyes, m-VoIP is not merely one of the numerous mobile apps but a seriously abnormal one. It could disrupt the revenue model that props up their licensed mobile service and network, on which majority of the users in the country rely on. The current fee for data usage is set relatively low because data service is bundled with voice service that is supposed to generate certain revenue, in which sense they do not think that the users’ payment for data usage fully justifies using m-VoIP. If m-VoIP is allowed freely to save payment for voice, such revenue

model can hardly be maintained and the resulting change in tariff structure could turn out to be worse for users as well as the telcos. So the current restriction on m-VoIP is not unfair and probably not even unfavorable to the users in the end, according to the telcos. And if they had to live in the era of unrestricted m-VoIP, they would want to charge network usage fees to the m-VoIP service providers, following the precedent of VoIP on broadband internet in Korea. So in terms of user benefits and fairness on the whole, it appears to be quite difficult to take one side over the other without conducting an in-depth investigation into such as alternative tariff structures and revenue models available to telcos. To narrow the focus on network neutrality issues, however, the telcos’ argument starts to lose its strength. Although m-VoIP problem must be among the most prominent issues in the discussion of mobile network neutrality, it is in fact a quite unique one distinguished from the others. In addressing wireless network neutrality in general, we usually try to be more careful considering the technological constraints of mobile network. Speed, capacity, and penetration of mobile broadband are typically much lower than those of fixed broadband and the existing mobile networks present operational constraints that fixed broadband networks do not typically encounter, which puts great pressure on the concept of reasonable network management for mobile providers (FCC, 2010). However, m-VoIP service itself does not cause much traffic, which means the telcos can hardly be convincing in justifying restriction of m-VoIP as reasonable network management. From the perspective of the telcos, P2P CDN and m-VoIP are almost the same kind of trouble makers in that they take away the existing source of revenues. And m-VoIP is in fact far more serious because the volume of the revenue threatened by m-VoIP must be huge. But, within the context of network neutrality, the telcos would not easily find appropriate measures to respond to the criticisms on their policy for m-VoIP. Moving a little beyond the area of network neutrality discussion, however, there is one potential

S. H. Kim / Communications & Convergence Review 3(1)

argument for necessity of charging fees to m-VoIP service providers. This argument (S. Kim, 2010) holds on the assumption that our society continues to demand a certain level of voice service provided by the traditional circuit switched network regardless of the success of m-VoIP service. If so, the circuit switched networks of the telcos could be regarded as close to public goods for which consumers are unwilling to pay in spite of their value to society. It implies that some fees may need to be charged for m-VoIP use to support the maintenance of circuit switched network, if it is not that government will take over the role of providing the circuit switched network. The most critical issue in the above argument is obviously whether the assumption can be justified. To my knowledge, the assumption will be convincing for now and many years to come. Circuit switched networks will not be completely replaced by IP networks for the reason of such as security and stability. Geer (2009) explains several factors that held back corporate m-VoIP adoption and threaten to continue doing so. Unreliable quality is among them. Because m-VoIP runs over the public internet, it does not provide predictable call connectivity or quality. The disruption of calls by bombarding the network with denial-of-service traffic is also a concern. And m-VoIP service can be affected even by weather conditions. Considering these technological limitations, users will want to keep the option of using circuitswitched voice service for such as emergency case or important business calls while most of the time using m-VoIP to save cost. Even an optimistic view that accounts for related technology advances predicts that 50 percent of all wireless traffic will be m-VoIP by 2019 (Geer, 2009), which means we would definitely need the service of circuit switched network at least for a decade to come. To summarize, m-VoIP case is a contentious issue in general but it reduces to a simple game that the telcos easily lose once network neutrality standard is introduced. It happens to reveal that the idea of network neutrality may be just one view stressing especially the value of using open network, which is more subjective than such values as user benefits or

97

fairness. Pursuing such value might be consistent with economic efficiency in short-term perspective where better utilization of network helps achieving static efficiency. However, it would not necessarily be so in longer term, where we should ensure incentives of all the players in the value chain of the industry to invest and innovate to achieve overall dynamic efficiency. Recognizing this leads us to look for some other policy beyond network neutrality. I explain it in the final section below.

Policy beyond network neutrality A policy beyond network neutrality is necessary to pursue the user benefits and the development of overall IP-based service industry in more universal and forward-looking perspectives. By universal perspectives, I intend to mean that openness may not be an ultimate goal of policy, especially in case that it conflicts with the user benefits. By forward-looking perspectives, I stress the importance of dynamic efficiency accounting for the incentive effects on all the players in the industry over time as well as user benefits in the long run. Then, such policy should expect and reflect various demands for IP-based services of the future. The network neutrality standards are limited because it hastily supposes that the existing class of platforms and networks will satisfy the future demands whose paths of development are in fact quite uncertain. D iscussion of the policy beyond netw ork neutrality can start by recognizing that the traditional internet sevice is now developing into 3 different IP service models, which are summarized in Table 2. The first type of them is the open network service like the current best-effort internet, which is supported especially by the end users. This kind of service will and should be maintained as we all appreciate enough the value of user choice and flexibility in open network. We cannot live without an open network any more since we know what it really can give us. The second type is the specialized service, which is provided by the platform or device providers. Smart-

98

S. H. Kim / Communications & Convergence Review 3(1)

Table 2. Development of 3 different IP service models Rules Applied Service Model

Major Player

Virtue

Internet Service

end users

Specialized Service Premium Service

Antitrust

Network Neutrality

Platform Neutrality

freedom flexibility

∨

∨

∨

platform or device providers

convenience flexibility

∨

∨

－

network providers

high quality

∨

－

－

phone, tablet PC, and smart TV services are the examples. This new type of services seems to improve the user convenience significantly by optimally integrating the service and the functions of device and platform. The third and last type is the premium service, which is supported by the network operators. Premium service here refers to the kind of service that capitalizes on the network technology integrating the functions of network, device, platform, and service altogether to produce the highest quality of IP services such as IPTV, N screen, video conferencing and so on. These services are not entirely separate as a user of one service can exchange data or content with one of another service when necessary, but they need to be recognized as distinguished from one another for their different operational or business models. It is a critical observation because the inherent differences in service models might require different rules and policies. A necessary assumption I make here is that the three service models are complementary, rather that substitutes, to one another. If, contrary to the assumption, users find that one service model is easily substituted for by another so that one economic market is defined for them, it would be preferred to have one policy to promote fair competition. The assumption of complementarity is justified because the types of IP service applications are extremely various and such variety is potentially unlimited so that one service model cannot support all of them well enough. Many applications have been working well on open internet, but it is not likely to be the same to all the applications developing now and in

the future. Such applications as location-based services, smart TV would be better served by specialized platforms due to their dependence on device functionality or security. A premium service guaranteeing a certain level of quality can be provided best by premium IP network service. So the three service models are not so much in competition to take a user or an application provider away from one another as in efforts to customize their platforms and networks to serve the specific needs of the users and application providers. In the eyes of the users and application providers, therefore, the three service models are complementary providing the differentiated options they can choose each time according to their specific needs. Some proponents of network neutrality may be opposed to the very idea of the IP service being divided. But, it is clearly different from the so called balkanized internet. Whereas the balkanized internet means the current internet being divided into some number of separate walled gardens, we consider the development of new types of services with different properties and virtues. Of course, if the new services end up replacing the current internet service, the result would be in effect the same as, or even worse than, what the opponents to balkanized internet are concerned about. Assuming complementarity of the service models, however, I do not concur with such concern. Still, I acknowledge it should be one of the main objectives of the policy beyond network neutrality to prevent the unintended result of losing open internet however little the possibility is.

99

S. H. Kim / Communications & Convergence Review 3(1)

And one should also notice that the development of different IP service models and policies is almost inevitable because one internet policy cannot satisfy different objectives that are pursued by different innovative players in the industry-internet users, platform or device providers, and network operators. Only the policy differentiated to different service models would enable the major players of each model to utilize their own specialized strengths for innovation. In trying to keep the open internet service intact, the key policy would of course be network neutrality policy, which includes ensuring the ongoing investment in the open network. But, one should not ignore platform neutrality, which requires that platforms do not block or discriminate specific applications or contents. Only when both network neutrality and platform neutrality are ensured, users can be allowed to fully enjoy the openness of the network. To be extreme, in the world of apps coming to substitute webs rather than complement them, users will have to concede some of their rights to choose to platform providers such as Apple or Google. The point is that we need to continue to have the internet where we are entirely free, where both network neutrality and platform neutrality exist. Even if I said that we need platform neutrality to ensure internet freedom, our preference need not be subject to one type of service model with its focus on freedom. That is why we are attracted to smartphones and tablet PCs. To benefit from this kind of specialized service, we need to accept some extent of control of platform providers as long as the control is appreciated to contribute to enhancing the service. It means we are willing to sacrifice platform neutrality to benefit from specialized service model. What about network neutrality? It is still an essential element for platform service providers doing business, evidenced by the fact that Google has been known for playing big role in promoting network neutrality. The premium service network is closed by its nature, which means neither network neutrality nor platform neutrality can be ensured, but it is not totally exclusionary to service providers. They can contract with and pay network usage fees to network operators

to provide service on the premium network. As for those relationships, whether they are real or potential, antitrust rules may be applied to ensure fair competition. While it is expected that the market of premium IP service will grow in the future, its expansion will consistently raise network neutrality issues as long as some fear that these premium services might replace competing services in the open network. To be optimistic, however, this premium service might be rather utilized to provide a valuable source of revenue that supports neutrality of the public internet. (It hinges on how the regulator makes effective arrangements with the network operators for the benefit of users.) Further, accommodating high bandwidth traffic in premium network could end up with helping the public internet maintain openness with less traffic burden. With this scenario of IP service development, the policy beyond network neutrality should look for an optimal balance of the three different IP service models. For example, if we pursue only a pure form of network neutrality policy, we might forego the chance of enjoying the premium service. In the opposite, if we ignore network neutrality policy, we might lose our valuable open internet service. So the reasonable policy should focus on providing appropriate incentives to invest in each of open network, specialized service, and premium service. The ultimate goal of pursuing the balance is to protect the user rights to choose and allow the users to benefit from using differentiated network services themselves. Allowing the consumers choose and pursue balanced bundles of consumption (given the assumption of convexity of preference) is one of the fundamental principles of economic theory.

References Geer, D. (2009). The Future of mobile VoIP in the enterprise. Computer, 42 (6), IEEE Computer Society. Ha, I., Wildman, S., & Bauer, J. (2010). P2P, CDNs, and hybrid networks: The economics of Internet video Distribution. International Telecommunications Policy Review, 17 (4), 1-22.

100

S. H. Kim / Communications & Convergence Review 3(1)

HSBC. (2010). Japan & Korea: Telecoms compared. HSBC Global Research. Kim, H. (2010). Suggestion for domestic network neutrality policy. KISDI Premium Report 10-09. Kim, S., Lee, N., & Kim, H. (2008). Network neutralityunderstanding its background and theoretical aspect. Korean Telecommunications Policy Review, 15(1), 95-133. Kim, S. (2010). In search of solution to network neutrality problems. DigiEco Focus. Kim, Y. (2010). Network neutrality for global leadership.

Presented in the symposium of Policies for Ensuring Global Leadership of Communications (October 15, 2010). KISDI Telecommunications Policy Research Division (2010). Subscriber numbers of telecommunication services. KISDI statistics (November, 2010). Lee, S., & Kim, J. (2008). A study on network neutrality regulation in information and telecommunication Industry. Policy Research Series 2008-12, KDI. United States. Federal Communications Commission. (2010). Report and Order, FCC 10-201.