Net Neutrality: An Untouchable Dogma?

Article  by  Mathieu PERONA  •  Published 28.01.2011  •  Updated 10.02.2011
Net neutrality movement symbol.
The principle of non-discrimination governs the Internet: on the Web, all data are treated in the same fashion. Will this be the case indefinitely?


On the information superhighway, liberty and equality reign supreme. No vehicle has the priority and the only speed limit is the one imposed by traffic; basic users have as many rights as the colossal Google or Facebook. Faced with congestion, the biggest and the most hurried users have to line up like everyone else – which brings about the debate opposing defenders and adversaries of net neutrality.

The net belongs to everyone: a norm under debate

Since the end of 2009, the term “network neutrality” (or “net neutrality”) has become a key element in any analysis of the telecommunications industry. The explosion of bandwidth-hungry services – such as music and video on demand – and mobile terminals has absorbed the excess capacity of the networks erected during the Internet bubble. Historically, the Internet operates on the basis of non-discrimination: all data are treated the same way, regardless of their source or destination.
As said Internet co-inventor Tim Berners-Lee in Scientific American article, this choice has profound implications on the operation of the Web, consisting of all the applications using the Internet: it means that none of these applications can be prioritized over another. Network neutrality will treat equally the data sent by individuals and businesses, data from games and from strategic communications, data from websites linked to the user’s ISP and data from other servers. For Tim Berners-Lee, this arrangement is at the basis of the Internet’s success, as it ensures that the Internet cannot be taken over by a company or an institution, and it makes all network resources equally accessible to individuals and innovative companies.
This defense of network neutrality relayed by regulatory bodies like the European Commission is opposed by the position taken by major network operators, first in line being Internet service providers, followed more recently by mobile network operators – unaccustomed to this philosophy – and players like Apple, which operate services reserved for owners of specific devices. For this group of stakeholders, the diversification of networks supporting data traffic (telephone landlines, mobile telephone networks, optical fiber, etc.), the rise of more bandwidth-intensive services (video conferencing, streaming, online games), and the multiplication of usage modes and media all require a way to differentiate data. There now exist tools capable of monitoring traffic in real time. Such tools make differentiated treatment technically possible, which may allow for additional sources of revenue to be reinvested in the funding of infrastructure development.
In this debate, economic analysis can identify several aspects of what net neutrality encompasses, and can assess the relevance of the arguments put forth by proponents of a differential data treatment for each of these aspects. As it stands, economic analysis alone is not enough to decide between these arguments. The available elements oppose a better use of existing networks with the fear of excessive content control by ISPs. It therefore does not support immediately imposing net neutrality throughout the sector. On the other hand, it calls for vigilance among consumers and competition regulatory bodies, who are going to have to closely examine how ISPs will seek to circumvent this de facto Internet standard.   
Back to summary

Why should we abandon net neutrality?

The arguments in favor of abandoning the principle of neutrality come mainly from Internet service providers (ISPs). Their two major angles are: 1) the willingness to use existing equipment more efficiently, and 2) the need to finance new infrastructures to cope with increased traffic. Indeed, current net traffic consists of content whose sensitivity to delays is highly variable, while deep packet inspection technology would give priority to the most sensitive streams and distribute the load more evenly throughout the day. However, these technologies will not address the increase in traffic from more intensive forms of network usage or the development of mobile platforms (smartphones, tablets). Therefore, the companies operating the infrastructure (for our purposes, referred to here as ISPs, although that term is not entirely accurate) are emphasizing the need to fund new equipment. However, network neutrality means that an ISP can only distribute these costs to their own customers instead of spreading them across the content providers who benefit from access to their customers. The central issue is that of financing equipment.
In light of these possibilities, network neutrality advocates point out that the Internet's success is based on the interconnection of all users. Besides the obvious possibilities of direct censorship, deep packet inspection would allow operators – often vertically integrated with content/service providers – to prioritize their own content offer to the detriment of competing content, as when mobile phone operators prohibit the use of VoIP protocols (like Skype) on their networks. The risk would be to fragment the Internet into ISP-controlled local territories.
In this debate, economic literature reveals that “net neutrality” actually covers two distinct regulatory elements: a non-pricing rule for inbound traffic on the one hand, and a no-discrimination rule on the other.
Back to summary


Network operators provide a platform between content providers (sites) and their subscribers. In broad terms, this kind of platform can draw revenues from both sides[+] NoteThe literature has coined the term “two-sided market” to describe markets where a platform puts in relation two sides of one market and where decisions taken on one side have a significant impact on the other.X [1] of its market by charging content providers for the opportunity to contact its subscribers, and by charging subscribers for connecting them to content providers. In its actual state, neutrality requires that operators do not make distinctions between incoming packets on their networks. This implies that ISPs cannot charge content providers for providing access to their sites. ISPs argue that the ability to charge large content providers such as Google and Facebook would allow them not to make their subscribers alone bear the costs of network development.
However, this option could have negative consequences. Since the early days of the Internet, most of the content is supplied by individuals who draw very little financial benefits, if any at all, and whose willingness to pay is thus very low. Even in the business world, companies now experiencing some success have gone through long periods when they were unable to transform an already high traffic flow into resources. Therefore, payment based on traffic is likely to impoverish content, form a barrier to new businesses’ Internet presence, and weaken the position of non-profit projects like Wikipedia and the Wikimedia Foundation’s other projects, which operate on donations.
Additionally, a coordination issue could complicate this impact. When an ISP requires a content producer to pay for access to its consumers, it benefits immediately from the payments of those who accept these terms, while the network’s loss of attractiveness – incurred from producers who prefer to throw in the towel – gets distributed over all ISPs through the consumer’s decreasing willingness to pay. Thus, even if Internet pricing could be useful, ISPs would tend to set rates too high, especially since competition in this field would be low, and users would lose their bargaining lever, i.e. switching ISPs. Ultimately, there would be a risk of balkanizing the network, with services accessible from some ISPs but not from others, at the expense of the network-wide effects that made the Internet successful.
In its current state, economic literature is very skeptical about the advisability of allowing network operators to charge their content providers for access to their network – especially since higher revenues would be much more to extend coverage (to gain new customers) than to improve existing infrastructures where subscribers are already captive, more or less.
Back to summary

Paying to reach Internet users

Net neutrality may also be seen as a non-discrimination obligation as to the origin or nature of transmitted content. Without such a requirement, operators could offer to content providers or users different contracts according to the type of bandwidth they want for each type of traffic. Alternately, operators could set priority rules without consulting anyone.
Since ISPs are hoping to force the companies benefiting most from the Internet (Google, Microsoft, Facebook, Blizzard and the like) to finance new infrastructure, the literature has mostly considered the “content provider billing” scenario. Content providers are unique because of their profit margins and their content’s sensitivity to eventual delay times. Depending on the extent of this distinction, with the introduction of a differential scheme that would provide priority access for a higher price, either all content providers would opt for priority access or a distinction would develop between priority access and slow access, the latter being chosen by content providers with lower profit margins or content insensitive to latency. In both cases, the network is used more efficiently, since delay times for the most valued or sensitive traffic are being reduced. The user experience would improve through the possible development of applications that can only operate on the basis of priority traffic, and content providers who opt for the slower service would have access to lower prices than in a single connection quality scenario. However, these contracts lead to the offer of a slower option than what can be offered and, compared to the single contract scenario, they transfer content provider profits to operators. Furthermore, contrary to their claims, it is unclear that such contracts would encourage them to invest in infrastructure. Increased network capacity mechanically improves slow-service quality, thus reducing the ability to draw higher added value from priority service. For this incentive to work, bandwidth improvement for both services would have to translate into a significant increase in consumer willingness to pay, which is not at all obvious in and of itself, especially in the context of a competitive individual access market.
In the end, breaking neutrality to introduce price discrimination would yield mixed results. It does not provide strong incentives for the development of the network, but it does allow a more efficient use of existing equipment, it would make services highly sensitive to delays possible, and it would transfer a portion of the most profitable content providers’ profits to network operators and access providers. The main risk would be a deterioration of basic service.
Back to summary

Defining priority rules

In the previous model, each content/service provider was selecting the plan (priority or slow) applicable to its data. However, it would be also possible to give operators the ability to prioritize traffic types they would consider priority traffic at their sole discretion, the assumption being that they are best placed to know how to efficiently use their capacities.
In fact, such practices already exist. In France, some ISPs restrict the flow of traffic associated with peer-to-peer protocols, while most operators on European mobile networks prohibit Voice over IP (VoIP). The latter case highlights the inherent dangers of such practices: most operators belong to groups that also provide services and therefore have an interest in degrading their competitors’ traffic, going so far as to exclude it altogether, as in Skype’s case.
As recalled by Neelie Kroes, European Commissioner responsible for new technologies, on 11 November 2010, consumer pressure warranted by the threat of switching ISPs is expected to punish unscrupulous operators (Kroes advised subscribers whose operators were prohibiting Skype to switch operators). However, competition in this market is doomed to decline now that most of the population is equipped and that substantial costs are incurred when switching operators. Therefore, we would have to trust businesses and competition regulatory bodies to put a stop to such practices when they are not justified by considerations of efficiency.
Back to summary

Risks on all sides

The economic literature does not allow us to rule in favor of any of the two positions. Abandoning net neutrality would allow a more efficient use of infrastructure pushed to their limits by usage growth, which would prevent the risk of a generalized congestion. However, it would favor current players, especially the biggest ones, at the expense of start-up businesses and non-commercial content providers. Meanwhile, consumers are torn between the benefit of cheaper subscriptions and the inconvenience of having to monitor their operator’s practices if they don’t wish to have access only to a subset of content. As it stands, the major issue is the vacuum in which this debate takes place: anti-competitive practices have been strongly contested and it is unclear whether operators themselves can benefit from the more harmful ones. The risk here becomes that of prohibiting legitimate practices in order to avoid possible deviations. From an economic perspective, it seems imperious to wait and use this opportunity to determine what importance should be given to each impact described above.
Back to summary


Crédit photo : Camilo Sanchez

Translated from the French by François Couture

Back to summary
  • 1. The literature has coined the term “two-sided market” to describe markets where a platform puts in relation two sides of one market and where decisions taken on one side have a significant impact on the other.
Would you like to add or correct something? Contact the editorial staff