Finding a Proxy for Net Neutrality

By: Ignacio Cofone, L.L.M. Candidate, Yale Law School. January 11, 2016.

In February 2015, the FCC approved net neutrality rules by classifying broadband as a utility under Title II of the Communications Act[i] and banning paid prioritization of data packages, which would effectively divide the Internet into “fast lanes” and “slow lanes.”[ii] However, in December, Republicans in Congress inserted a rider into the appropriations bill that would stymy the FCC’s ability to enforce net neutrality.[iii]

Net neutrality can have at least two different meanings. In its broad conception, net neutrality is a prohibition on interfering in any way with the content delivery networks between some companies and ISPs. Violations of this concept of neutrality are common, and it is difficult to ensure perfect net neutrality in this broad sense. In a narrower sense, it means preventing ISPs from blocking or prioritizing traffic.[iv] Violations of this concept of neutrality only became possible decades after the Internet was created,[v] but banning this practice is perfectly feasible.

Several countries around the world are debating whether they should preserve the way the Internet has functioned so far, along the standards of narrow net neutrality, or allow ISPs to block traffic and prioritize some bits and clients over others.[vi] In Europe, the issue has been debated since 2011, and this past October, the European Parliament voted in favor of Europe-wide net neutrality.[vii] In India, there is ongoing controversy. While the Information Technology Act does not establish net neutrality,[viii] the Telecom Regulatory Authority of India supports the idea, but does not enforce it. The government has called for comments and suggestions from the public—a response to them is still forthcoming.

One way to approach the question of whether we should have net neutrality in a narrow sense is to ask ourselves whether people have a right to access the Internet. Such a right could exist on its own, or it could derive from the right to access information. If there is such a right, then net neutrality is necessary to allow people to access the Internet more freely and more widely and to prevent barriers such as increased prices from unfairly limiting usage.

If such a right does not exist, however, then the choice between net neutrality and non-neutrality is a matter of governmental discretion, and decisions about how to regulate are best approached using a framework that evaluates consequences, rather than principles.[ix] The obstacle to evaluating consequences in this case, it is sometimes argued, is that so far there is no country in the world with non-neutrality, so it is difficult to anticipate exactly what will happen, and we can at best make conjectures. This, however, is only partially true, since we can try to find a proxy.

Contrary to what one might think, China has a system similar to non-neutrality. Although ISPs in China do not engage in packet filtering or make decisions about what content is allowed or blocked based on market pressures,[x] there is a version of one centralized authority choosing and prioritizing certain content for all. That one authority—the State—does so through four state-owned Internet backbones[xi] that control traffic from ISPs, which must be licensed.[xii]

It seems, therefore, that the difference between the Internet situation in China and the Internet situation that could result in the West if non-neutrality prevails is, in essence, that in China the choice is made by a governmental monopoly, while in the West the choice would be made by companies competing in the market place (or an oligopoly).[xiii] This is not so much a difference in kind as a difference in degree.

Examining China’s form of non-neutrality allows us to make more precise conjectures. The question becomes, what aspects of the Internet situation in China (which the West generally considers undesirable[xiv]) could be prevented if non-neutrality were implemented, not via a monopoly, but via an oligopoly or large companies in perfect market competition.

Non-neutrality might present a benefit. With net neutrality, ISPs argue that they have fewer incentives to invest in research and development. Using two-sided market models, some have argued along these lines that non-neutrality is a good idea because it allows for price discrimination, where everyone pays his or her reservation price, thereby eliminating social waste and, allegedly, increasing incentives to invest. But the literature on whether net neutrality actually decreases ISPs’ incentives to invest, as ISPs argue, is divided.[xv] At an abstract level, this is connected to an ongoing discussion in industrial organization of whether companies have more incentives to invest in a monopoly (because they have extra profits to do so) or in perfect competition (because they have pressure to improve their products to beat their competitors and maintain or increase their market share).[xvi] The benefits of non-neutrality are less clear than how they are often presented.

On the other side of the scale, unlike in China, the market for ISPs in the West might provide some hope. If ISPs compete, then users will be able to avoid ideologically biased ISPs in favor of more neutral alternatives, so ISPs would have an incentive to maintain their neutrality.

However, this “market correction” would only take place in the scenario of a market without failures, where the choice of an ideological (conservative-funded, liberal-funded, libertarian-funded, etc.) or non-ideological ISP is as easy as a choice between red and green apples. Many markets in our society are like that, but many others are not due to a lack of transparency or complexity in the products.[xvii] Still, even these latter markets are often left without intervention because such intervention would be more costly than the welfare enhancement of correcting the failure (if such correction is even possible). 

Thus, we might question whether the market would work so smoothly. The market would be able to filter out ideological ISPs only when there is not an adverse selection problem created by the difficulty of screening for ideologically driven ISPs. An adverse selection problem occurs when there is asymmetric information about a product, and the party with less information (often a consumer) cannot properly screen between high and low quality products. This will lead her to decide on a product and a price regardless of the quality (which she cannot assess), even if it will negatively affect her later on.[xviii] If it is too difficult for consumers to know which ISPs filter packages ideologically and to what degree, they will choose one either ignoring this dimension or giving it less weight than they would if meaningfully informed—both forms of market failure. In turn, the correction of the market failure is already in place. Net neutrality fixes the problem of the inability to screen ideology in ISPs by forbidding them to be ideological altogether. It operates similarly to other regulations that approach adverse selection problems, such as minimum quality standards and mandatory warranties, by eliminating low quality in the dimension in which screening is difficult or impossible. Moreover, the potential loss of eliminating this particular regulation is substantial: it risks not only consumers’ dissatisfaction, but also, due to the nature of the product, a reduction in their civil liberties.

Once we dispel the myth of achieving an absolutely level playing field (net neutrality in the broad sense), we see that net neutrality is somewhat similar to how we treat other things in the offline world. For each consumer, it operates like many other regulations that protect her from products that she cannot properly screen. For the broader societal consequences, that an absolutely level playing field is impossible is not reason enough to allow for harmful differentiation. Discrimination is only the most obvious example. Often, we prefer to make the playing field as level as we can, even if it will never be perfect.

Our scarce available experience with non-neutrality and the conjectures we can make about it seem to indicate that our current system of net neutrality is a good idea, and probably the best path. Good ideas have a tendency to bring together people with different points of departure, who might separately miss a part of the big picture. In terms of net neutrality, reasonable people who think there is no right to access the Internet, but who wish to limit the impact of ideological ISPs, agree in their policy conclusion with other reasonable people who think that such a right does exist. Indeed, perhaps we should think twice, or thrice, about revoking net neutrality.

[i] Communications Act of 1934, 47 U.S.C. § 202.

[ii] Dana Liebelson, Net Neutrality Prevails In Historic FCC Vote, Huffington Post (Feb. 26, 2015, 1:02AM), neutrality-fcc-vote_n_6761702.html[].

[iii] The rider states: “None of the funds made available by this Act may be used to regulate, directly or indirectly, the prices, other fees, or data caps and allowances (as such terms are described in paragraph 164 of the Report and Order on Remand, Declaratory Ruling, and Order in the matter of protecting and promoting the open Internet, adopted by the Federal Communications Commission on February 26, 2015.” Appropriations for Fiscal Year 2016,[http://]. See also Amanda Terkel, Provision Killing Net Neutrality Threatens Must-Pass Spending Bill, Huffington Post (Dec. 1, 2015, 5:17PM), neutrality-omnibus_565e0303e4b08e945fecf41d?mq0bpgb9 [].

[iv] Joshua S. Gans, Weak Versus Strong Net Neutrality, 47 J. Reg. Econ. 183 (2014).

[v] ISPs have only been able to do this since 1999, when Cisco released the first router that could inspect data packages and allow the network operator to prioritize, de-prioritize, or block them.

[vi] The issue gained prominence in the U.S. last year when Netflix paid Comcast for a fast lane, generating controversy.

[vii] EU Actions: Our Commitment to Net Neutrality, Eur. Commission (Oct. 27, 2015, 3:56PM), []. Some have argued, however, that the net neutrality norm is undermined by several loopholes. see Alex Hern,EU Net Neutrality Laws Fatally Undermined By Loopholes, Critics Say, The Guardian (Oct. 27, 2015), [].

[viii] Information Technology Act, 2000 (Law No. 21/2000). The law is based on UNCITRAL’s Law on Electronic Commerce model of 1996.

[ix] This does not necessarily mean performing a utilitarian analysis or a strict calculus for the maximization of some other metric, but generally evaluating the possible consequences of the measure and deciding whether we find them appealing.

[x] Jack L Qiu., Virtual Censorship in China: Keeping the Gate Between the Cyberspaces, 4 Int’l J. Comm. Law and Pol’y  1 (1999). The Chinese government also engages in other practices to control Internet content, such as DNS filtering.

[xi] Bryce T. McIntyre, China's Use of the Internet: A Revolution on Hold, in Telecommunications and Development in China (Paul S.N. Lee ed., 1997).

[xii] Temporary Regulation for the Management of Computer Information Network International Connection, of 1996. The intermediaries are currently China Telecom, China Golden Bridge Network, China Education and Research Network, and China Science and Technology Network. There is also a prohibition on broadcasting news and media from outside of China without government approval, established by State Council Order No. 292 of 2000.

[xiii] One could also say that the government operates similarly to an oligopoly in China, and then both situations would be more alike.

[xiv] The situation in China has widely been referred to as “censorship,” see, e.g., Beina Xu,Media Censorship in China, Council on Foreign Rel.(Apr. 7, 2015), [],and as a society, we arguably try to avoid censorship.

[xv] For arguments that net neutrality does diminish incentives to invest, see Nicholas Economides and Benjamin Hermalin, The Economics of Network Neutrality, 43 RAND J. Econ. 602 (2012); Paul Njoroge et al., Investment in Two-Sided Markets and the Net Neutrality Debate, 12 Rev. Network Econ. 355 (2013). For arguments that net neutrality does not harm ISPs’ incentives to invest, see J.P. Choi and B.C. Kim, Net Neutrality and Investment Incentives, 41 RAND J. Econ. 446 (2010).

[xvi] See Andreu Mas-Colell, Michael D. Whinston & Jerry R. Green, Microeconomic Theory 383-414 (1995).

[xvii] This would take place when people cannot tell the difference between products, as the next paragraph shows.

[xviii] George Akelrof, The Market for Lemons: Quality, Uncertainty and the Market Mechanism, 84 Q. J. Econ. 488 (1970). In an extreme version of the problem, where consumers are unable to screen any dimension of the good, the market will unravel.  See also Andreu Mas-Colell, Michael Whinston & Jerry Green, supra note 16, at 436-50.