Mode 1, Mode 2, or Mode 10: How Should Internet Services Be Classified in the Global Agreement on Trade in Service?
By: Usman Ahmed, Brian Bieron, and Gary Horlick
In the case of United States–Measures Affecting the Cross-Border Supply of Gambling and Betting Services (US – Gambling), the Appellate Body of the World Trade Organization treated an Internet-based service as a Mode 1 service – a cross border service supplied within the territory of the member. The Internet is more complex than this explanation, though, with elements of both Mode 1 and Mode 2 services (i.e with consumers entering the market of the service provider and requesting the service). This classification is a threshold legal issue in trade law, as it determines nations’ obligations under the General Agreement on Trade in Services (GATS), and it has arguably been misinterpreted for the last two decades. This piece focuses on the technical aspects of the Internet, particularly its “push” and “pull” protocols, and seeks to clarify how the Internet operates in order to aid future cross-border electronic trade cases. The interpretation of which Mode covers Internet services has significant implications for trade law. Thus, it is essential to understand the technology in order to make sound legal judgments.
Laws and regulations often struggle to keep up with innovations in technology given both the speed of technological development and the lack of adequate analogies for policymakers to use in applying old laws to new technologies. Trade law is no different in this respect. As technology rapidly changes the global economy, trade law will continue to struggle to appropriately address regulatory barriers that prevent the widest application of the most groundbreaking advancements in technology. The Internet is a perfect example of this phenomenon.
In 1995, the National Science Foundation divested its final piece of its computer science network (NSFNET), marking the official commercialization of the Internet. At the time, the Internet had a user base of less than two million. A year earlier, the World Trade Organization (WTO) had been created to “to develop an integrated, more viable and durable multilateral trading system.” The General Agreement on Trade in Services (GATS) was signed as a part of the negotiations with the aim of removing regulatory barriers to services trade.
In the mid-1990s, people could only speculate as to how the Internet would come to affect global trade law. Now, in 2015, over three billion people use the Internet. The McKinsey Global Institute reported that the Internet accounts for 21% of GDP growth in mature economies, with the majority of the economic gains accruing to traditional firms. Moreover, small and medium sized enterprises use this technology to directly trade with consumers in other countries — a stark shift in the nature of global trade. The US International Trade Commission found that “[i]n 2012, online sales by digitally intensive small and medium-sized enterprises (SMEs) were $227.1 billion, or about one-fourth of total online sales, and online purchases by SMEs were $162.2 billion, or about one-third of total online purchases.”
The WTO has made efforts to tackle the issues of the Internet over the past two decades, through its Electronic Commerce Work Programme and most notably through the Appellate Body’s ruling in the 2005 case United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services (US — Gambling). The US — Gambling case presents an example of the difficulties associated with attempting to govern new technologies using classical legal doctrine.
This paper seeks to clarify and criticize a particular finding made in US — Gambling in an effort to better fit the issues of the Internet in trade law going forward. The paper will proceed in four parts: Part II will analyze US — Gambling and its treatment of the Internet. Part III will describe alternatives to the analysis used in US — Gambling proposed both by governments and academic literature. Part IV will explain how the Internet operates from a technical perspective and suggest a new method for analyzing the GATS in light of the architecture of the Internet. Finally, Part V will discuss the implications of how Internet services are analyzed under the GATS.
II. The US – Gambling Case
The island nation of Antigua and Barbuda (“Antigua”) is home to several online gambling websites. These gambling businesses reach customers around the world and transact with them over the Internet. According to Antigua, 105 US federal and state measures, three judicial cases, and several government actions have prohibited the type of cross border gambling services that Antiguan businesses were seeking to provide to US consumers.
In 2003, Antigua brought a claim before the WTO arguing that the United States had violated its commitment to provide market access for cross-border recreational services under the GATS by imposing a set of federal and state laws that barred online gambling services. WTO’s decision was groundbreaking as it was the “first [case] within the WTO framework to examine cross-border electronic trade in services—the vehicle for outsourcing many types of services.” Moreover, it confirmed that WTO rules would apply to Internet services. The threshold question in the case was whether the United States had made commitments under the GATS to provide market access for “recreational services,” including online gambling services. To determine whether the US had made such a commitment, the WTO panel first needed to determine the “Mode” of supply of an Internet service. The GATS outlines four Modes:
Mode 1: from the territory of one member to the territory of another;
Mode 2: in the territory of one member to service a consumer of another country;
Mode 3: through an established commercial presence in the territory of another country;
Mode 4: through the presence of a natural person.
Nations have agreed under a schedule of commitments to provide, or not provide, market access to specific services (such as recreational services) based upon their Mode of delivery. Electronic services could presumably be supplied to another country through any of the four Modes, but there has been a longstanding disagreement in the scholarship and among nations over whether Internet based services are delivered as Mode 1 or Mode 2. The question is whether the Internet service is providing the service across borders (Mode 1) or a consumer is “traveling” abroad in order to access an Internet service (Mode 2).
Several meetings of the WTO Electronic Commerce Work Programme discussed the issue of how to classify Internet services, but failed to reach a consensus. Thus, it was up to the WTO’s judicial arm, the Appellate Body, to determine how to treat Internet gambling services. The Appellate Body decision did not discuss whether Internet services were Mode 1 or Mode 2, but instead relied upon the findings of the panel. The panel implied that Mode 1 applied in the case of online gambling because online gambling services were delivered remotely, a characteristic exclusive to Mode 1 services. The decision promotes a commonly held assumption that Mode 1 is the exception to the rule that there must be physical proximity between the service provider and the consumer. The panel’s treatment of the issue, though, was limited and there were no definitive statements on the Mode of supply; it could not be said to be dispositive on the issue. Notably, the decision of whether the online gambling services were Mode 1 or 2 was irrelevant to the outcome of US — Gambling, as the US had similar market access commitments to “recreational services” under its Mode 1 and Mode 2 obligations. Nevertheless, following the publication of the case, government officials and scholars accepted that Internet services would henceforth be governed as Mode 1 services. This paper seeks to challenge that longstanding notion.
III. Rethinking the Interpretation of Internet Services Under the GATS
It is worthwhile to examine alternatives to the approach used in US — Gambling that have been proposed over the years before delving into the ultimate conclusion of this paper. A number of papers were published in the early 2000s that sought to settle whether Internet services should be classified as Mode 1 or Mode 2, largely in response to disagreements in meetings of the WTO E-Commerce Working Programme. This section will review the gamut of solutions proposed in the literature. The first is perhaps the simplest.
1. Classifying Internet Services as Mode 2
The United States proposed that Internet services be classified as Mode 2 services during the meetings of the WTO E-Commerce Working Programme and in the US — Gambling case itself. The rationale behind a Mode 2 classification was that it would align with the general agreement to open up market access, or liberalize. This classification would have the benefit of clarifying that it is a consumer decision that leads to accessing an Internet service, rather than a service provider initiating the transaction. However, this interpretation would have the same problem as the existing approach: namely, it would overlook important technical aspects of the Internet, which are discussed in detail below.
2. Combining Mode 1 and Mode 2 into a Single Category
Combining the two Modes would eliminate any classification confusion because there would no longer be any difference between Mode 1 and Mode 2. However, there are important distinctions between Mode 1 and Mode 2 services which are worth preserving. A consumer visiting another country to take advantage of a service raises different jurisdictional and judicial concerns than a service provider does in entering the consumer’s country to deliver a service. Moreover, nations negotiating the GATS made differing commitments for Mode 1 and Mode 2 services, with the understanding that they referred to different practices. The distinction between Mode 1 and Mode 2 has practical significance and should be maintained going forward.
3. Creating a new “Mode 5” or “Mode 10” to Encompass Internet Services
This interpretation would attempt to recognize the unique nature of the Internet and avoid unintended modal classification difficulties by creating a new Mode for Internet services, which could be called “Mode 5” or “Mode 10.” This new Mode could also foster the creation of a digital sector for the purposes of market access commitments and a digital chapter in future trade agreements. However, it is unclear how this new Mode would be defined. Moreover, as the Modes are defined in the GATS, the creation of a new Mode would require a renegotiation of the entire treaty, likely making this solution untenable. As will be discussed in further detail below, Internet services can, in fact, be classified as either Mode 1 or Mode 2 services and, thus, it is unclear if a “Mode 5” or “Mode 10” would add any value to the GATS.
4. Making Mode 1 and Mode 2 Commitments Identical
The distinction between Mode 1 and Mode 2 was not central to the US — Gambling decision because the United States’ commitments for “recreational services” under Modes 1 and 2 were the same. Matching Mode 1 and Mode 2 commitments across all sectors would eliminate the problem of modal classification would be eliminated by. However, as mentioned above and described in greater detail below, there are important jurisdictional and substantive differences between Mode 1 and Mode 2, nations have made commitments relying upon these distinctions, and it would require a large undertaking to completely reengineer GATS commitments for all sectors.
5. Clarifying that Mode 2 Services Require a Physical Presence in the Jurisdiction of the Consumer
This interpretation would provide a clear distinction between Mode 1 and Mode 2 services. It would eliminate the need for large-scale renegotiation of existing commitments, and it seems that many of the members signing onto the GATS thought of Mode 2 services in this manner. The confusion between Mode 1 and Mode 2 exists beyond Internet services, and this interpretation has the benefit of resolving the question for all sectors. This would potentially limit Mode 2, though, beyond the scope of what was originally intended.
For example, if a business in France picks up the phone and requests that a Canadian architectural firm fax updated blueprints to Paris, would we say this is a Mode 1 service? The business in France has essentially entered Canada to consume the architectural services, but has not established any physical presence in Canada. However, if classified as Mode 2, this would mean that the business in Canada had entered the territory of France to deliver services, which seems inaccurate. This potential physical presence requirement of Mode 2 should not be confused with the commercial presence requirement of Mode 3.
6. Determining Liberalization Based upon Sector
A rather radical solution would be to review all sectors and set aside some as “sensitive” areas (e.g. financial services or healthcare) which, when delivered over the Internet, ought to be viewed as Mode 1 services with less market access liberalization. Others sectors that are less sensitive (e.g. software) would be deemed Mode 2 services when delivered over the Internet, with greater market access liberalization. This solution, though, would require a complete overhaul of the GATS system to carve out sensitive sectors. Moreover, there would be tremendous difficulty in determining which sectors ought to be considered sensitive. Finally, the modal classification issue exists beyond Internet services, and limiting this solution to Internet-based services would prove problematic from an interpretation perspective.
7. Enabling Each Member to Decide How it Classifies Internet Services
The concern that members are not aware of what types of services they have liberalized could be resolved by deferring to members to determine whether an Internet service is Mode 1 or Mode 2. Uncertainty over classification can provide members with a justification for avoiding commitments altogether. This solution would be untenable, though, because the lack of uniformity in understanding the nature of the modes would lead to confusion and unpredictability in areas beyond Internet services.
All of the solutions described above would raise troubling issues of interpretation, feasibility, unknowable future implications, conflicts with existing commitments and infrastructure, or lack of technical accuracy. The solution proposed in the following sections attempts to resolve these concerns.
IV. How the Internet Works
The Internet has traditionally utilized a “pull” based model for disseminating information. Pull technology requires a user to enter a query before a service provider responds (e.g. a Google search). Users, rather than service providers, initiate a “pull” by explicitly requesting information in a pull technology enabled communication. “Push” technology, by contrast, sends information out to the user, typically after the user has signed up for a service and specified what he or she wants to receive (e.g. real-time news updates on the BBC news app). Push technology sends data to users without them having to actively request the data. Service providers, rather than users, initiate the “push.” Radio, television, and telephones can all be thought of as forms of push technology.
The Internet can be used for both push communications (e.g. e-mail) and pull communications (e.g. the typical website).  Push technology, however, has traditionally been seen as intruding upon a user’s web experience and therefore accounted for only a minority of Internet traffic. That has begun to change, however, as today’s mobile smartphone technology has developed to allow for more push information than ever before, particularly for users that download apps and consent to receive automated real-time updates. This is true to such an extent that the phrase “push notification,” signifying a piece of information sent from the service provider to the user without an explicit user request, is now common parlance for smartphone users.
An interim report of the Electronic Commerce Work Programme noted that Internet services can appear to function as either Mode 1 or Mode 2. This is true, as information delivered as a push is more like Mode 1, and information delivered as a pull is more like Mode 2. It is unfortunate that neither the members of the Work Programme nor scholars have picked up on the distinction between push and pull technology, as this distinction could provide a clear differentiation between when an Internet service is being delivered as Mode 1 or Mode 2: when information is delivered as a push it could be considered Mode 1 (from the territory of one member state to the territory of another), and when it is delivered as a pull it could be deemed Mode 2 (in the territory of one member state to service a citizen of another member state).
Scholars and governments have expressed confusion about “where” a consumer and service provider meet in cyberspace for the purposes of modal classification. Our proposed classification based upon push and pull technology should help scholars, countries, and courts to interpret model classification for the Internet going forward.
V. The Implications
Classifying an Internet services as Mode 1 or Mode 2 has implications for trade commitments, jurisdiction, and domestic regulatory policy. For example, over 80% of nations have agreed to open up communications services by Mode 2, but less than 20% of agreed to open up Mode 1. Conversely, only 25% of nations have agreed to open up business services by Mode 1, while less than 15% have agreed to open up Mode 2. In general though, nations have agreed to open up market access for Mode 2 to a greater extent than they have for Mode 1.
However, Mode 2 is not necessarily better for business than Mode 1. Only Mode 1 requires that the capital movements related to the underlying transaction be free of restrictions. This is an important liberalization, as financial transactions often underlie cross border Internet transfers and a lack of protection for these underlying transactions under Mode 2 would certainly be more restrictive for business. Thus, from a trade commitment perspective, there are pros and cons to being deemed either a Mode 1 or Mode 2 service.
Jurisdiction is a threshold issue that is determined when deciding whether an Internet service is Mode 1 or Mode 2, as it determines which nation’s laws apply to a particular dispute. If Internet services are deemed to be Mode 1 services, then the national legal system of the consumer would apply because the service provider is supplying the service into the nation of the consumer. If Internet services are deemed to be Mode 2 services, then the legal system of the service provider would apply because it is the consumer that is entering the market of the service provider to obtain the service.
Scholars have argued, therefore, that a Mode 1 interpretation promotes the notion of consumer protection because a nation can challenge a foreign service provider without presence, while a Mode 2 interpretation promotes global trade because foreign service providers are free to provide services despite domestic regulations. The simplified regulatory compliance burdens inherent in Mode 2 are a more pronounced factor for very small enterprises, which should be an important consideration given the expanding use of Internet-enabled services by small and medium-sized enterprises to engage in cross-border commerce.
Upon closer inspection, though, this tradeoff between consumer protection and global trade appears to be illusory. Mutual recognition standards, best practices, and international agreements are all methods by which consumer protection can be achieved without requiring a service provider using the Internet to subject itself to the jurisdiction of every country where its website or application could be accessed, and the nature of the global Internet lends itself to these types of cooperative solutions. Even if an Internet service is viewed as Mode 2, there are still domestic regulations that nations can impose on domestic Internet Service Providers (“ISPs”) in an effort to combat notions that they feel are problematic. The domestic ISPs solution is certainly less than ideal, as it cuts off an Internet service completely, but it does uphold the principle of jurisdiction that requires some minimum contacts in order to exercise judicial authority over an entity.
The most significant impact of the Internet from an economic perspective has been the ability of a small business to rapidly scale around the world. Nearly every business can become an exporter, and nearly every consumer can become an importer. The economic benefits of this development mean a more positive, inclusive economy for smaller businesses and lower prices for consumers. Governments ought to consider the benefits of Internet services when determining the question of modal interpretation, as uncertainty over the classification of Internet services under the GATS could limit these benefits. If a small business could be held liable in every jurisdiction from which its website could be accessed, then the small business may be deterred from launching websites.
The issue of trade and Internet services is only going to grow in importance as the number of people and enterprises connected to the Internet increases. Due to increasing access to and use of the Internet, services that were once considered “untradeable” (e.g. architecture, health, education) are now increasingly tradable due to the Internet. There are a host of market access barriers to Internet services in the form of discriminatory licensing regimes, foreign investment caps, and state owned enterprise set-asides. Nations’ filtering and blocking of Internet services has been a major issue for the last twenty years, and will remain so going forward. It is imperative that liberalized service trade commitments be reached to address these problems and enable the free flow of Internet services. The correct treatment of Internet services in trade policies can only help spur the development of the Internet economy as investors and businesses become more confident in the to policy regime.
There has been a longstanding disagreement in trade law circles as to whether Internet services should be deemed as Mode 1 or Mode 2 services. A host of solutions have been proposed by governments, lawyers, and academics. This paper has suggested a novel solution to the modal interpretation problem which mirrors the technical architecture of the Internet: when an Internet service is pulled by the user, it could be deemed to be a Mode 2 service, and when it is pushed to the user, it could be deemed to be a Mode 1 service. This solution would not require an overhaul of the GATS and would provide a clear distinction between the Modes. Moreover, this solution would neither require a renegotiation of commitments nor the tackling of difficult jurisdictional questions. As Internet services continue to proliferate to every service sector, it will become increasingly important to resolve this modal interpretation issue. We hope that this paper can help in that effort.
 Usman Ahmed is Head of Global Public Policy at PayPal. Mr. Ahmed has a B.A. from the University of Maryland and a J.D. from the University of Michigan. Brian Bieron is the Executive Director of eBay Inc.’s Global Public Policy Lab. Mr. Bieron has a B.A., magna cum laude, from St. Joseph’s University and a M.A. in International Affairs from George Washington University. Gary Horlick is an international trade lawyer based in Washington DC. Mr. Horlick received his A.B. from Dartmouth, his B.A. & M.A. from University of Cambridge, and his J.D. from Yale. Together, the three authors teach a class on the Internet and International Trade Law at Georgetown University. This piece solely reflects the views of the authors in their academic capacity and does not represent the views of any company or client.
 A Brief History of NSF and the Internet, National Science Foundation, http://www.nsf.gov/news/special_reports/cyber/internet.jsp (last visited Oct 30, 2015).
 Marrakesh Agreement Establishing the World Trade Organization, Apr. 15, 1994, 1867 U.N.T.S. 154.
 General Agreement on Trade in Services, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1B, 1869 U.N.T.S. 183 [hereinafter GATS].
 See, e.g., The Internet Economy: The World’s Next Growth Engine, Business Week (Oct. 4, 1999), http://www.businessweek.com/1999/99_40/b3649004.htm (“But with the Internet, it becomes much easier to provide services of all types–banking, education, consulting, retailing, gambling–through a Web site that is globally accessible. . . . Indeed, the real key to the next stage of the Internet Age will be whether financial innovation diffuses as fast as technological innovation.”)
 Internet used by 3.2 Billion People in 2015, BBC News (May 26, 2015), http://www.bbc.com/news/technology-32884867.
 Jacques Bughin et al., The Macroeconomic Impact of the Internet, in Internet Matters, Essays in Digital Transformation 13 (McKinsey Global Institute, 2012).
 U.S. Int’l Trade Comm’n, Pub. No. 4485, Digital Trade in the U.S. and Global Economies, Part 2, 14 (2014).
 See infra Part II.
 See, e.g., Bodog, www.bodog.ag (last visited Nov. 1, 2015); Sports Betting, www.sportsbetting.ag (last visited Nov. 1, 2015).
 Joseph M. Kelly, et al., Clash in the Caribbean: Antigua and U. S. Dispute Internet Gambling and GATS An interview with Joseph M. Kelly, 10 UNLV Gaming Res. & Rev. J. 15, 17 (2006).
 Appellate Body Report, United States – Measures Affecting the Cross‑Border Supply of Gambling and Betting Services, WTO Doc. WT/DS285/AB/R (adopted Apr. 20 2005); Sacha Wunsch-Vincent, The Internet, Cross-Border Trade in Services, and the GATS: Lessons from US–Gambling, 5 World Trade Review 319, 321 (2006).
 Joel P. Trachtman, United States—Measures Affecting the Cross-Border Supplying of Gambling and Betting Services, 99 Am. J. Int’l L. 861, 862 (2005)
 Appellate Body Report, supra note 13, at 323.
 Tim Wu, The World Trade Law of Censorship and Internet Filtering, 7 Chi. J. Int’l L. 263, 277-78 (2006).
 GATS, supra note 5, at Part I, Art. I.
 Judson O. Berkey, A Framework Agreement for Electronic Commerce Regulation Under the GATS 1-2 (Institute of International Finance 2001) (describing classification as a threshold issue); Emad Tinawi & Judson O. Berkey, E-Services and the WTO: The Adequacy of the GATS Classification Framework 3 (Organization of Economic Cooperation and Development 1999).
 Dora S. Neo, China, India and Global outsourcing of services under GATS, in China, India and the International Economic Order (Muthucumaraswamy Sornarajah, Jiangyu Wang eds., Cambridge University Press 2010); Sacha Wunsch-Vincent, WTO, E-Commerce and Information Technologies: From the Uruguay Round to the Doha Development Agenda 85 (Joann McIntosh ed. 2004); Daewon Choi & Solange Ghernaouti-Helie, An Analytical point of view of emerging issues of e-commerce for the multilateral trading system 14 (2003), http://kavehh.com/my%20Document/WTO/an%20analytical%20point%20of%20view%20of%20emergin%20issues%20for%20r-commerce%20for%20the%20multilateral%20trading%20ssytem.pdf.
 See generally Berkey, supra note 18. See also Aaditya Mattoo & Ludger Schuknecht, Trade Policies for Electronic Commerce 14 (World Bank Policy Research, Working Paper 2380, 2000); Heinz Hauser & Sacha Wunsch-Vincent, A Call for a WTO E-commerce Initiative, 6 Int’l J. of Comm. Law and Policy 33 (Winter 2000); Pascale Bourquin & Zora Wilkinson, Committee on the General Agreement on Trade in Services, Oikos Model WTO 14 (2014).
 Tinawi &Berkey, supra note 18, at 3.
 Wunsch-Vincent, supra note 13, at 320; Neo, supra note 19, at 287.
 See Appellate Body Report, supra note 13, at 81 (“[W]e therefore, uphold the panel’s finding that: Please provide a citation a citationo Therese mentioned to me as Thoreau seems to suggest) does civil disobedience become a jus[a prohibition on one, several or all means of delivery cross-border] is a “limitation on the number of service suppliers in the form of numerical quotas” within the meaning of Article XVI:2(a) because it totally prevents the use by service suppliers of one, several or all means of delivery that are included in mode 1.
 Wunsch-Vincent, supra note 13, at 326; Wu, supra note 16, at 278.
 Tinawi & Berkey, supra note 18, at 2.
 Arvind Panagariya, E-Commerce, E-Commerce, WTO and Developing Countries, in Policy Issues in International Trade and Commodities Study Series No. 2 16 (United Nations 2000).
 Wunsch-Vincent, supra note 13, at 326.
 Wunsch-Vincent, supra note 13, at 325.
 United Nations Conference on Trade & Development, A Positive Agenda for Developing Countries: Issues for Future Trade Negotiations 463 (2000).
 Tinawi & Berkey, supra note 18 at 9; Bourquin & Wilkinson, supra note 20, at 15; Andrew D. Mitchell, Towards Compatibility: The Future of Electronic Commerce Within the Global Trading System, 4 J. of Int’l Econ. Law, 683, 690 (2001).
 See infra Part V.
 Mitchell, supra note 31, at 691-92; Bourquin & Zora, supra note 20, at 15.
 Panagariya, supra note 27, at 17-18.
 Wunsch-Vincent, supra note 13, at 326.
 Bourquin & Wilkinson, supra note 20, at 15.
 Mitchell, supra note 31, at 691-92; Leon Steenkamp, Challenges Posed by Electronic Commerce to the Operation and Implementation of the General Agreement on Trade in Services 10 (Tralac, WP 2/2004, 2004).
 Mitchell, supra note 31, at 690.
 Ernesto Hernández-López, Trade in Electronic Commerce Services under the WTO, 4 J. of World Intell. Prop. 557, 574 (2001).
 Rolf H. Weber, International E-Trade, 41 Int’l Lawyer, 845, 857 (2007).
 Hernández-López, supra note 39, at 575.
 Michael Vax, The Internet is About to Take a U-turn—Is Your Business Prepared?, The Future of Customer Engagement and Commerce (May, 18, 2015), http://www.the-future-of-commerce.com/2015/05/18/the-internet-is-about-to-go-from-pull-to-push-is-your-business-prepared/.
 Julie E. Kendall & Kenneth E. Kendall, Web Pull and Push Technologies in Emerging Information Technology: Improving Decisions, Cooperation, and Infrastructure, 265 (1999).
 Michael Franklin & Stan Zdonik, “Data in Your Face”: Push Technology in Perspective, ACM SIGMOD 516 (1998), http://ftp.cs.toronto.edu/cs/ftp/public_html/pub/hfliu/papers/XFilterVLDB00.pdf.
 Kenneth W. Umbach, What is “Push Technology”?, 4 Ca RB Note 6, 9 (1997).
 Franklin & Zdonik, supra note 44.
 Id. at 516.
 Umbach, supra note 45, at 3.
 Id. at 516-17
 See Dmitry Namiot & Manfred Sneps-Sneppe, Conference on Future Internet Communications, Local Messages for Smartphones 1-6 (2013).
 Tinawi & Berkey, supra note 18, at 5.
 Somkiat Tangkitvanich, Global E-commerce Policies seen from the South, in Getting Connected: Information and Communications Technology for Development 25 (Cosmas Gitta, ed. 2001).
 Id. at 6.
 Wunsch-Vincent, supra note 13, at 324; Weber, supra note 40, at 856..
 Steenkamp supra note 37, at 9.
 Mattoo & Schuknecht, supra note 20, at 18; Bourquin & Wilkinson, supra note 20, at 14.
 Wunsch-Vincent, supra note 19, at 86.
 Id.; Hauser & Wunsch-Vincent, supra note 20, at 10.
 Mitchell, supra note 24, at 691; Tangkitvanich, supra note 40, at 25.
 Berkey, supra note 18, at 4.
 Mitchell, supra note 31, at 691.
 Dan L. Burk, Jurisdiction in a World Without Borders, 1 Va. J. L. & tech 3, 26 (1999).
 Pierre Sauvé, The Trade Policy Implications of the New Economy, Center for International Development at Harvard University E-Commerce Papers 4 (2001), www.cid.harvard.edu/cidtrade/Papers/Sauve/sauvetpimp.pdf. See also Hauser & Wunsch-Vincent, supra note 20, at 2 (describing welfare increasing characteristics of e-commerce).
 Wu, supra note 12, at 267.
 Usman Ahmed & Brian Bieron, Services, International Rulemaking, and the Digitization of Global Commerce, International Centre for Trade and Sustainable Development 9 (Apr. 16, 2015), http://www.ictsd.org/themes/global-economic-governance/research/services-international-rulemaking-and-the-digitization-of.
 Weber, supra note 32, at 857.
 Berkey, supra note 18, at 2.
 See Aimee Chanthadavong, Zukerberg, Bill Gates Pledges for Universal Internet Access by 2020, ZDNet (Sept. 28, 2015), http://www.zdnet.com/article/zuckerberg-bill-gates-pledges-for-universal-internet-access-by-2020/.
 Sacha Wunsch, Electronic Services: Its Regulatory Barriers and the Role of the WTO, Center for International Development at Harvard University E-Commerce Papers 6 (2001), www.cid.harvard.edu/cidtrade/papers/wunsch
 Hauser & Wunsch-Vincent, supra note 20, at 5.
 Panagariya, supra note 27, at 19.
 Sauvé, supra note 65, at 6; Mark J. F. Lund & Steven McGuire , Institutions and Development: Electronic Commerce and Economic Growth, 26 Org. Stud. 1743, 1757 (2005).