Cornerstone Course – Day 5: Digital Society – Network Neutrality

Digital Society is a very elastic phrase. We will explore three examples:

  1. Network Neutrality
  2. Privacy and Surveillance
  3. Big Data

All are focused on how technology changes society. It is a contested topic on whether the impact is positive or negative. Issues are at the intersection of information and communications technologies and society, law, and public policy.

Network Neutrality

The term dates back to the telephone line networks. It is focused on the Internet, but there are more kind of networks which rely on the same principle.

The main idea is that traffic over the network should all be treated the same. No package should be singled out for its user, content, origin or destination.

“The principle that Internet service providers and governments should treat all data of the Internet equally, not discriminating or charging differentially by user, content, site, platform, application, type of attached equipment, or mode of communication” – FCC

The Wallstreet journal said in 1990 that the major effect of the Internet would be the replacement of the fax machine – which would be a major change. However, it eventually failed at that as the fax machine provides a receipt for having sent and printed a file which the US legal system recognize as a serving a subpoena.

Minitel was a France initiative to look up phone numbers on a small computer in the 1980. In the follow-up it offers weather forecasts and soon allowed private companies to offer services. It was quite successful, however, it was always run by France Télécom S.A. (top-down) and eventually the Internet overtook it because the internet was bottom-up.

In 2008 2/3 of the Internet bandwidth was used by BitTorrent. The share was actually decreasing. However, the FCC was investigating Comcast whether they slowed BitTorrent connections. This was the precedence for the FCC to enforce net neutrality. The next round was fought between Netflix and Comcast in 2012 when Netflix accused Comcast of throttling their video services to favour their own. Out of court settlements were reached and end-user speed on Netflix increased. Another case happened in 2015 when the Marriott hotels blocked wireless spectrum to prevent customers from operating their own mobile Wi-Fi hotspots. After a public outcry they retreated from the policy.

In Switzerland communication providers offer their own video services without limitation (Salt/Zatttoo and Swisscom/Swisscom TV Air).

Networks and layers

Networks in general are built in layers. Specifically upper layers built on lower layers.

[table id=6 /]

The Internet is only fixed in the IP protocol, technology above and below can be changed. This is called the Internet hourglass model. The network types on the bottom of the glass and the applications above at the top of the glass have evolved tremendously.

On the downside IP is very constrained. Improving it is very difficult, but denial of service attacks (DOS), security concerns and the limited number of addresses require changes.

Network neutrality has a technical foundation that it requires the end-to-end communication to work. However, there are two perspectives.

  1. Functions should only be implemented in a lower layer if it can be completely & correctly implemented at that layer
  2. In addition, if the function is needed by all clients of that layer

On the one hand, the advantages are that the network doesn’t know anything and doesn’t do things that are better done at the end. The argument naturally emerges from the technical structure of the internet. Additionally, it provides the following advantages:

  • Long-term evolvability
  • Application autonomy
  • Reliability
  • Minimizes interfaces between modules
  • reduces complexity

Those advantages are claims and not necessarily correct. On the other hand, disadvantage is that the internet does not do things well, but it does them all. Phone calls for instance. Phone lines were optimized for verbal communications. When a fax machine was connected it had to realize that, stop echo cancelling, stop compressing. The phone system typically ran at 90% utilization whereas the internet rarely runs above 20%.

Carrying public internet traffic is not economically viable as the traffic does only cost but cannot provide income. Sprint subsidized the public internet by phone systems and today the internet is subsidized by large companies buying private networks.

Modularity as a design principle offers advantages by minimizing interfaces between modules and reducing complexity. Facebook started out as a website out of a dorm room. However, the design tends to be static and can have negative effects on

Virtual Private Networks (VPNs) and Tunnels allow to disguise the content of your communication (e.g. human right activists hiding communication from a repressive government or drug cartels organising their deliveries). This poses a problem for law-making if they want to change net neutrality. Encrypted content cannot be identified and therefore the loss of net neutrality would most likely result in the illegality of VPNs.


The US has different legal framework for DSL and cable modem providers. Telecommunication services and information services have different legal requirements. Telecommunication services are more restricted and enforced net neutrality whereas information services do not. DSL is a telecommunication services and cable modems are information services. The legal definitions are not clear and still contested in the courts. The FCC tries to reclassify cable modems to be telecommunication services.


  • Does a neutral network discriminate against Quality of Service applications? (e.g. Skype)
  • Does the argument hold in a world of competition between DSL and cable?
  • Do we need special regulations? Isn’t this topic for antitrust policy?
  • Is this a debate about economic effects or about freedom?
  • Situation in Europe and Switzerland?
  • Technical solutions to the network neutrality problem? How to resolve this transparently?