On Tuesday, October 27, the European Parliament will vote on rules intended to protect network neutrality in the European Union (EU). However, the proposal about to be adopted fails to deliver network neutrality to the EU and is much weaker than current net neutrality rules in the United States. Fortunately, it’s not too late to change course. Members of Parliament can still secure meaningful network neutrality for Europe – if they adopt key amendments on Tuesday.
This is a guest post by Barbara van Schewick, which first appeared at Medium. Barbara van Schewick is a leading net neutrality expert, Professor at Stanford Law School, Director of the Stanford Center for Internet and Society, and the author of Internet Architecture and Innovation (MIT Press 2010, Paperback 2012). Her writings on network neutrality have influenced regulatory debates in the United States, Canada, Europe, Latin America and India. The FCC’s Open Internet Orders in 2010 and 2015 relied heavily on her work. Parts of this text draw on her earlier writing on net neutrality.
Unless it adopts amendments, the European Parliament’s net neutrality vote next Tuesday threatens the open Internet in Europe.
The European Parliament understands that the future of economic growth, innovation, and free speech in Europe depends on net neutrality – the principle that has kept the Internet an open and free space since its inception. However, a compromise proposal up for vote next Tuesday contains major problems that threaten the open Internet in Europe. Contrary to some claims, the proposal is weaker than network neutrality rules in the US. European citizens deserve the same free and open Internet that Americans can enjoy.
The good news is that members of the European Parliament will introduce amendments that would fix these problems. For the amendments to be adopted, the majority of the members (376 of the 751 members) need to vote for the amendments.
The current proposal has major problems.
The proposal bans Internet service providers (ISPs) from blocking or slowing down websites, or charging sites extra fees to reach people faster, i.e. creating “fast lanes” online. This is good – companies that provide gateways to the Internet shouldn’t interfere with our ability to access what we want online.
But the current proposal contains four significant problems that still allow ISPs to engage in bad behavior that would harm Internet users, businesses, and speakers in Europe.
- Problem #1: The proposal allows ISPs to create fast lanes for companies that pay through the specialized services exception.
- Problem #2: The proposal generally allows zero-rating and gives regulators very limited ability to police it, leaving users and companies without protection against all but the most egregious cases of favoritism.
- Problem #3: The proposal allows class-based discrimination, i.e. ISPs can define classes and speed up or slow down traffic in those classes even if there is no congestion.
- Problem #4: The proposal allows ISPs to prevent “impending” congestion. That makes it easier for them to slow down traffic anytime, not just during times of actual congestion.
We should ask the Parliament to adopt amendments to ensure an open Internet in Europe.
To save the open Internet in Europe, members of the European Parliament need to adopt the amendments. This won’t happen automatically. Here’s what you can do to help:
Take action: Ask your representatives in the Parliament to adopt the necessary amendments. You can find all the necessary information and tools at SavetheInternet.eu.
Spread the word: Share this post and others on Facebook, Twitter, or anywhere else. Talk with your friends, colleagues, and family and ask them to take action. If you are a blogger or journalist, write about what is going on.
HOW THE AMENDMENTS WILL FIX THE PROPOSAL
Let’s look more closely at the major problems with the compromise proposal and how to fix them.
The proposal allows ISPs to create fast lanes for companies that pay through the specialized services exception.
“Fast Lanes” on the Internet harm innovation, free expression, and democratic discourse in Europe.
ISPs want the power to charge websites extra fees to reach people faster. If some websites can pay ISPs to be in the “fast lane,” anyone who can’t afford the extra fees will be left behind in the slow lane. That means that it will be harder for Europeans to access websites that can’t afford to pay extra fees. European start-ups, small businesses, non-profits, educators, artists, musicians, writers, activists, faith groups, and NGOs would be at a disadvantage. Europe’s Internet would no longer be an open space where everyone has an equal chance of reaching people.
The proposal allows ISPs to create harmful fast lanes online.
The current proposal generally bans ISPs from creating fast lanes. But it contains an exception for “specialized services.” That is important, because it allows applications to emerge that would not be able to function on the open Internet because they need special treatment that the open Internet cannot provide.
However, the exception for specialized services is too broad: In many cases, it still allows ISPs to offer fast lanes by calling them a specialized service.
Fast lanes would crush start-up innovation in Europe and make it harder for European start-ups to challenge dominant American companies.
Slow lanes spell disaster for innovation. On the Internet, the costs of innovation are incredibly low. Entrepreneurs don’t need permission to innovate or a lot of upfront funding. This benefits European start-ups that often don’t have access to outside funding.
Allowing ISPs to offer fast lanes would change that. If established companies can pay so that their content loads faster or does not count against users’ monthly bandwidth caps, then those who can’t pay don’t have a chance to compete. Many of today’s most popular applications – Google, Facebook, Yahoo, eBay – were developed by innovators with little or no outside funding. In a world where such innovations are stuck in the slow lane, they would never have seen the light of day. Today, these companies can pay for a fast lane, but the next generation of innovators that will challenge them should be assured the same level playing field.
Fast lanes would harm all sectors of the economy.
Today, most companies rely on the Internet to reach their customers. As a result, fast lanes would harm all sectors of the economy. Large corporations that pay to be in the fast lane will have higher costs, so we the customers will be forced to pay higher prices for their products and services. Small businesses that are unable to pay will be shut out of the market. Many small businesses in Europe are just starting to take advantages of the opportunities that the Internet has to offer. They should have the same opportunity to benefit from an Open Internet as their counterparts in the United States.
SOLUTION: The Parliament should adopt amendments that refine the definition of specialized services to close the specialized services loophole and keep the Internet an open platform and level playing field.
The proposal generally allows zero-rating and gives regulators very limited ability to police it, leaving users and companies without protection against all but the most egregious cases.
Zero-rating is harmful discrimination.
Zero-rating is the practice of not counting certain applications against users’ monthly bandwidth caps. Like fast lanes or other technical discrimination, zero-rating allows ISPs to discriminate against content that users want to see. Zero-rated applications are more attractive to users than applications that are not. That means zero-rating some content over others has the same discriminatory effect as speeding up certain applications over others. Like fast lanes, zero-rating threatens the future of competition, innovation, and free speech in Europe and around the world.
The proposal does not clearly ban zero-rating.
The European Commission and the European Parliament disagree whether the proposal applies to zero-rating. The Commission has stated that zero-rating is subject to the rules and that the rules allow it. If zero-rating is subject to the rules, regulators’ ability to regulate zero-rating is limited: regulators could only reach the most egregious cases, while everything else would be permitted. Member states would not be able to adopt additional rules to regulate zero-rating – their hands would be tied.
However, members of the Parliament said that they agreed with the Council during the compromise negotiations that zero-rating would be outside the scope of the rules.
Zero-rating results in discrimination.
Research shows that zero-rated applications are far more attractive to users than those that are not. In a study commissioned by the CTIA, 74% of users said that they would be more likely to watch videos offered by a new provider if the content did not count against their monthly bandwidth caps. When the online magazine Slate experimented with zero-rating, it told some users that the podcast did not count against their cap. People who were offered the zero-rated podcast were 61% more likely to click on the link. Thus, zero-rating has the same impact as technical discrimination: it gives ISPs power to make certain applications more attractive than others and pick winners and losers on the Internet.
Zero-rating distorts competition.
In the European Union, many ISPs zero-rate their own video applications. Users on these plans can watch unlimited zero-rated videos, but their bandwidth caps prevent them from watching more than 2-5 hours of video content unaffiliated with the ISPs. Similarly, many ISPs in Europe zero-rate their own cloud-storage applications. Their users can upload 10 gigabytes of traffic to the ISP’s cloud storage for free. But it costs between $50 and $70 to upload the same amount of data to other cloud storage sites like Dropbox or Google Drive. These plans make it effectively impossible for unaffiliated providers to compete with the ISP’s zero-rated application.
Zero-rating harms users.
When European ISPs start zero-rating certain applications or content, they often reduce overall bandwidth caps or increase the price of unrestricted bandwidth, as the European research firm Rewheel has shown. This is not surprising: the lower the bandwidth caps, the more attractive zero-rated applications become, so lower bandwidth caps motivate rich providers to pay for zero-rating. Thus, zero-rating harms users (and the providers of applications that are not zero-rated) by reducing the amount or increasing the costs of bandwidth that users can use however they like. By contrast, when a Dutch regulator banned zero-rating, the provider KPN doubled its monthly bandwidth cap for mobile Internet access from 5 to 10 GB at no additional cost.
Zero-rating harms innovation and free speech.
Start-ups, small businesses, and low-cost speakers in Europe and elsewhere don’t have money to pay for fast lanes; they don’t have money to pay for zero-rating, either. But if some companies can pay to be zero-rated, those who can’t pay will find it hard to compete. Thus, allowing ISPs to zero-rate websites or services against a fee creates the same problems for innovation and free speech as allowing ISPs to charge for fast lanes.
SOLUTION: Europeans deserve clarity on this important issue. The parliament should adopt amendments that make it clear that member states are free to adopt additional rules to regulate zero-rating. This would bring the proposal in line with the negotiators’ stated intent and empower individual member states to address this harmful practice in the future.
The proposal allows ISPs to define classes and speed up or slow down traffic in those classes, even if there is no congestion. This allows ISPs to distort competition, stifles innovation, harms users, and hurts providers who encrypt traffic by putting all encrypted traffic in the slow lane.
The proposal allows ISPs to engage in class-based discrimination.
The proposal allows class-based discrimination: ISPs can make distinctions between different kinds of traffic and treat them differently to optimize overall transmission quality at any time, not just during times of congestion. The discrimination must be based on the technical requirements of the applications in question. Thus, ISPs could treat different kinds of applications differently if they have different technical requirements. For example, Internet telephony is sensitive to delay, but e-mail is not, so an ISP could give low delay to Internet telephony, but not to e-mail.
Whenever an ISP has the power to speed up certain applications or slow down others, it might use this power to give certain applications an advantage over others. The proposal tries to mitigate this danger by forcing ISPs to consider an application’s technical requirements when making distinctions among traffic.
However, this kind of class-based discriminatory network management still allows ISPs to give some applications an advantage over others, whether intentionally or inadvertently. It distorts competition, slows all encrypted traffic, harms individual users, stifles innovation, and creates high costs of regulation.
Allowing ISPs to treat classes differently gives them power to deliberately distort competition.
When ISPs are free to define classes, they have a lot of discretion to discriminate against certain applications. ISPs could use this power to deliberately distort competition. For example, an ISP could offer low delay to online gaming to make it more attractive, but it could decide not to offer low delay to online telephony because that would allow Internet telephony to better compete with the ISP’s own telephony offerings. Although both services are sensitive to delay, ISPs could argue that there are other, technical differences that justify distinguishing between them.
Class-based traffic management can inadvertently harm applications.
Traffic management that distinguishes among different kinds of applications often results in inadvertent discrimination that hurts users, distorts competition, and makes it harder for providers of affected applications to innovate. Traffic management technologies that distinguish among classes of applications often end up harming certain applications, even if that effect is not intended, because the ISPs or their technology misclassify certain applications.
For example, many ISPs in the UK limit the bandwidth available to peer-to-peer file sharing applications during times of congestion, arguing that these applications are not sensitive to delay. This creates huge problems for online gaming. ISPs use deep packet inspection technology to identify these applications, but the technology doesn’t work very well: it has a hard time distinguishing between online gaming and peer-to-peer file sharing, so online games stop working or don’t work as well as they could. In the end, UK ISPs and gaming providers established standing committees where ISPs, technology vendors, and gaming providers worked together to make sure the games would work on ISPs’ networks in spite of the discriminatory network management.
In the UK, this class-based traffic management not only creates problems for online gamers and gaming providers, whose applications perform worse than other kinds of applications, but it also creates problems for innovation. If an online gaming provider wants to introduce a new feature for its game in the UK, it needs to work with the ISPs and their technology vendors to make sure that the feature won’t be caught up in the traffic management measures directed at peer-to-peer file sharing. This is the opposite of innovation without permission.
Similarly, until 2010, many ISPs in Canada used deep packet inspection technology to single out all peer-to-peer file sharing applications and limit the amount of bandwidth available to them from 5pm to midnight. Again, ISPs assumed that it was alright to target peer-to-peer file sharing, because it’s not sensitive to delay. But this assumption turned out to be wrong: there was an application called Vuze that used peer-to-peer file sharing protocols to stream video in real time. Real-time video is highly sensitive to delay, so the performance of Vuze suffered in the evening, when everybody wants to use the Internet.
Thus, the class-based traffic management might result in harmful discrimination by even the best-intentioned ISPs.
Class-based traffic management discriminates against encrypted traffic.
If traffic is encrypted, then the ISP cannot identify what kind of application – e-mail, telephony, web browsing – that a user is using, so it doesn’t know what kind of treatment it needs. In the past, ISPs have addressed the problem by simply putting all encrypted traffic in the slow lane. That means that any time someone sends encrypted data, it will take longer to transmit. People encrypt their data for a variety of valid reasons, for example, to protect privacy, secure sensitive financial transactions, protect trade secrets, and guard against surveillance. If all encrypted data is automatically slowed down, it would discourage people from using encryption at all.
Class-based traffic management harms individual users.
Class-based traffic management takes the power to choose the right kind of service out of the hands of users and puts it into the hands of ISPs. However, people have different needs for speed on the Internet, and the same person has different needs at different times. As a result, a user’s needs may differ from an application’s technical requirements, so ISPs don’t necessarily know what kind of service a user needs. For example, Internet telephony applications like Skype benefit from low delay, so ISPs may opt to give them low-delay service. That’s great if you are doing a job interview, where you want the best quality possible. But if you are talking with a friend, you don’t need crystal clear quality over Skype, so low-delay service might not be necessary. File uploads are generally considered not to be sensitive to delay. If you are uploading your hard disk to the cloud to do a backup, you will not mind that ISPs give file uploads lower priority. But if you are a student uploading homework right before it’s due, or a lawyer filing a brief before the deadline, or an architect submitting a bid, then the speed of this upload is your highest priority. As long as ISPs, and not users, have the power to decide which classes of application get what kinds of service, users will never get exactly what they need. That’s why class-based discrimination often harms users.
Class-based traffic management stifles innovation.
Imagine you develop a new application that would benefit from a specific kind of service. Entrepreneurs and start-ups typically do not have the resources or capacity to reach out to ISPs around the European Union to alert them that their particular application needs a certain kind of service. Even if a start-up manages to contact ISPs, they may not be interested in changing their systems for particular applications, which is a lot of work, especially when new apps don’t have any users yet. Entrepreneurs should be able to get the kind of Internet service their application needs without having to seek ISPs’ permission.
Class-based traffic management leads to high costs of regulation.
If ISPs get to define classes of applications, the only way to challenge these definitions is to complain to regulatory agencies. The agency would need to determine whether kinds of traffic are similar enough to be treated in the same way, a messy and costly process that would involve lots of lawyers and expert witnesses. This not only creates high costs of regulation, but also tilts the playing field against anybody – users, start-ups, small businesses, low-cost speakers – who doesn’t have the money to engage in long and costly proceedings before a regulator.
SOLUTION: The Parliament should adopt amendments that prohibit ISPs from differentiating among classes of applications unless it’s necessary to manage congestion or maintain the security or integrity of the network. The proposal already contains an exception that allows ISPs to differentiate among classes of applications if that’s the only way to manage congestion. Thus, removing the language that allows class-based traffic management at all times protects users, competition, and innovation, while still giving ISPs the tools they need to manage their networks.
The proposal allows ISPs to start managing congestion in the case of impending congestion. That means that they can slow down traffic anytime, not just during times of actual congestion.
The proposal gives ISPs a carefully calibrated set of tools to manage congestion. In particular, it allows them to differentiate among classes of traffic if doing so is necessary to deal with temporary or exceptional congestion. The relevant provisions have been designed carefully to ensure that discriminatory traffic management remains the exception rather than the rule.
However, the proposal allows ISPs to use the same tools to prevent “impending” congestion. Since the meaning of “impending” is not clearly defined, this provision opens the floodgates for managing traffic at all times. It makes it easier for ISPs to discriminate among classes of applications even if there is no congestion, using the justification that congestion was just about to materialize.
SOLUTION: The Parliament should adopt amendments that limit ISPs’ ability to discriminate against traffic to prevent impending congestion.
HOW WE GOT HERE – AND WHAT WE SHOULD DO NEXT
In April 2014, the European Parliament voted for strong network neutrality rules for the European Union. But under European law, network neutrality rules need to be adopted jointly by the Parliament and the Council, which consists of the representatives of the governments of each member state. Over the past year, the Council has consistently supported proposals that were significantly weaker than the Parliament’s text. In late June, representatives of the Parliament and the Council unexpectedly reached a compromise in informal negotiations. This compromise proposal was formally adopted by the Council in September.
If a majority of the members who vote approves this flawed compromise next Tuesday, the rules are adopted and become law. Europe will have far weaker network neutrality rules than the US, and the European Internet would become less free and less open. By contrast, if a majority of the members approves amendments, the text goes back to the Council. The Council can then accept the amendments, and they become law. If the Council rejects the amendments, a joint committee consisting of representatives of the Parliament and the Council has six weeks to come up with a compromise. Any compromise would then have to be adopted by the Parliament and the Council.
The future of the Internet in Europe is on the line. It’s up to all of us to save it.