2015 11 11 WS 197 Net Neutrality: Yes, No or Maybe? Workshop Room 9 FINISHED

The following are the outputs of the real-time captioning taken during the Tenth Annual Meeting of the Internet Governance Forum (IGF) in João Pessoa, Brazil, from 10 to 13 November 2015. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the event, but should not be treated as an authoritative record. 

 

***

>> MODERATOR:  Hello.  Welcome to the panel Net Neutrality:  Yes, No or Maybe. 

And we have lots of ‑‑ this is IGF and there are lots of panels on net neutrality, but I think this is the best because we have all the rock star speakers here and we have a different format. 

So it won't be boring.  I can assure you it will be a fun panel, and they will discuss.  And we have like two or three camps:  Yes, No, and Maybe. 

And we tried to come up with a question for the panel, but then we couldn't really find the question. 

That's why we left it ambiguous because we say the question is net neutrality; and we have like three camps saying yes, no, or maybe. 

So we have ‑‑ like, as I told, you we have rock star speakers here. 

On my left side, it's Berin Szoka, Christopher Yoo, Mike Nelson, Bob Pepper, Eduardo Bertoni; and last, but not least, Harold Feld. 

Okay.  We have a hashtag as well ‑‑ it's net neutrality ‑‑ for your tweets.  And let's start.

As I told you before, we couldn't come up with a question; but before we start, I want to ask the audience like who are in the yes camp for net neutrality? 

I see. 

Three, four, maybe eight ‑‑ no, nine.  Ten.  Okay.  I'm writing this down because at the end of the panel, I'm going to ask you again.  And who says no? 

Six no. 

Okay.  We have six people in no camp. 

And who are the maybes? 

Yes. 

(Discussion off microphone)

>> MODERATOR:  Yes, I think that should be the question, who is not interested in this debate. 

So how many do we have?  14?  Okay. 

>> BERIN SZOKA:  And before we get started, let me make clear that our moderator here is going to look at Twitter and look at any questions that you want to ask to inform the debate.  So we're not going to have questions from the floor because this topic has been thoroughly discussed, and we want to leave time for more Oxford‑style debate.  So we'll go from the yes camp to the no camp to the maybe camp, do it one more time; and then we're going to ask each other questions, interrogatories; and then our moderator will ask whatever questions we haven't answered or that she takes from the floor that you think are important.  And then we'll have closing statements.

And Mike is going to need to leave a little early.  And then at the end we'll take a vote and see if anybody actually changed their minds.

>> AUDIENCE:  (Off microphone)

>> MODERATOR:  Okay.  So shall we start with Harold.

You have six minutes, and I will be strict time keeper. 

>> HAROLD FELD:  Thank you. 

I start with the proposition that the question of network neutrality and the use of specific rules prohibiting blocking, prohibiting paid prioritization, prohibiting degradation of content and access spring from a fundamental human right to communicate.

We have in the Internet the communications medium of the 21st century; and for one of the first times in human history, it is possible to communicate globally potentially without the means of an intermediary, that we do not have necessarily the licensee or the company that controls the pipe to control this means of communication. 

We had this in the 20th century with the telephone, but the telephone was limited to one‑person‑to‑one‑person audio communication. 

In the world of mass media which shapes opinion, provides so much content and influence to people, we traditionally had mediation.  We had a few broadcasters.  We had a few people that chose what programming goes on the satellite or on the cable. 

The Internet has changed that, and it's the neutrality of that network that we borrow from the phone net, applied to the technology of the Internet which makes all manner of communication possible, including one to many, to be a fundamental change in the way in which people communicate and which brings about what many had considered an impossible ideal of genuine communication. 

This principle of network neutrality with specific rules not only protects the individual and the individual rights, it protects the carrier as well.

We have seen too often soft censorship, efforts to censor websites such as WikiLeaks or other unpopular speech or the speech of minorities or even speech that people regard as distasteful such as hate speech; but we have seen the efforts of governments to circumvent their judicial processes and impose restraints on the third parties that carry the speech, not direct legal restraints subject to judicial review but instead through the soft power of promising favors to those companies that comply with the government's requests and promising disfavor to those who do not.

To have laws which make it impossible for the carriers to comply with such requests shields carriers from this judgment.  It, likewise, shields carriers from the judgments of the market with regard to unpopular speech. 

The short version of my position is that where a carrier will claim editorial power, it will be asked to exercise that power.  Where the carrier cannot exercise editorial power over speech, the fundamental human right of free communication is protected throughout the network except for direct government intrusion which can be labeled for what it is, censorship, or subject appropriate to the appropriate due process restraints. 

Finally, I will add that while I have stressed this fundamental human right as being the core of why these rules are necessary, I would add that they also turned out to be very good for business, that in a world that is increasingly complicated, we find that while it is theoretically possible that there may be forms of discrimination that are favorable, history shows that the transaction costs are very high. 

In many places, there is not choice of provider.  And then as a matter of economics, as well as a matter of fundamental human rights, settled expectations with regard to the behavior of the carriers so that investments can be made at the edge and in core so consumers will know what they can receive. 

So fundamental rights can be protected and are best served by network neutrality with specific rules to protect against adverse conduct. 

I yield back the remainder of my time. 

>> MODERATOR:  Thank you very much. 

So the next is Christopher from the no camp. 

>> CHRISTOPHER YOO:  Thank you, all, for coming. 

At a less abstract level, I took the debate of the program literally as should governments impose specific ex ante roles to govern network neutrality.  The way the question focuses there are two ways to go, set up rules in advance or have a case‑by‑case, after‑the‑fact assessment.  And, in fact, those are the true traditional modes that are dominant, that you have in the discourse of network neutrality. 

Interestingly, this is the way we've enforced almost all legal principles over the years.  And there's a mature set of scholarship that looks at when you do one and when you do the other. 

You use rules in advance when you're familiar with the practice and it's very clear that by ruling out a certain kind of conduct, there is no harm or that conduct that you're ruling out is so rarely beneficial it's not worth looking at the details of how it works. 

You use after‑the‑fact review on a case‑by‑case basis when the phenomenon you're dealing with is new and complex and not well understood and you don't know whether the benefits will, in fact, outweigh the cost or vice versa. 

I'll give you an example.  I'm a parent ‑‑ this is how I teach my students this principle.  The children ask, "When do I have to be home tonight?"  

I say, "11:00.  Well, you have to be in the house." 

"What if there's a fire in the house," or "What if it's my birthday" or a special occasion. 

And we do all the exceptions and you don't understand it very well.  And, oh, you have children.  My son is a future lawyer.  He will argue with me forever.

The other example, on a more substantive level, is speed limits.  We now have a rule for speed limits, 55 miles an hour or 70 kilometers an hour.  In the original days of cars, we did not have that.  We said, "Drive safely under the circumstances."

Now, you cannot determine in advance what that conduct is; but you see as you put an obligation on a person to be reasonable in light of all the circumstances, that is a classic stance. 

So the very question of net neutrality is which is the more appropriate way to effectuate net neutrality, rules in advance or creating a standard and exploring it on case‑by‑case basis.

So I love having this talk at IGF because the IGF is the perfect forum to show just how complex or how new practices on the Internet really are. 

One of the things we found out in the news is we found out it's a different reality the way these network neutrality principles play out in the developed world verses the developing world.

In the developed world, the big news in the United States is the number four carrier, T‑Mobile, has done a zero‑rating program for music called Music Freedom.  All streaming music does not count against the data cap. 

Yesterday, they announced they're going to do the same for streaming video for 24 different services including Netflix, HBO, ESPN, some of the most popular services in the United States.

What you see is this is the number four carrier who is attempting to differentiate the services they are offering; otherwise, they have to compete on price and network quality which are dimensions that are going to favor the incumbent.  So this is an attempt by a small player to compete and to give people a reason to buy. 

There's an excellent example of MetroPCS, the number eight carrier, did something similar before and was subject ‑‑ and was able to do it on 1G quality spectrum and was sued for net neutrality violation.

So now we understand there are many forms of deployment even in the developed world.

And I'm doing studies about building Native American communities in the U.S., some hyperworld strategies in the U.S., and they're all doing innovative strategies in this regard which are nonneutral.

In the developing world, we're seeing the debate emerging over zero rating; and we understand when there is no service, it is a different question about the way these policies play out.  And, in fact, in other places where you have a single monopoly service, there's still a different context that needs to play out.

We need to understand this is a complex phenomenon with different ways.  And, in fact, as fact as Harold notes, the early sessions on net neutrality at this conference have said even if you frame it in terms of human rights, there are times that you have to put limits on what carriers do and there's times you have to put obligations on what they do because there are competing rights that are at issue. 

So what we're seeing is it's a very complex phenomenon that requires a careful balance. 

The other part of the point I'd like to make is not a competitive one.  It's about innovation.  The problem with ex ante rules is that innovation needs room to breathe and experiment. 

The default answer should be yes unless you show form of harm. 

Ex ante have the opposite effect.  The default answer is no unless you go to a regulator and convince them to authorize the process. 

I'm worried that the great innovation world that we have in the Internet where the default answer is yes will turn around because if you have an innovative new practice that conflicts with the one of the old rules, you cannot do it unless you go into a political process which will represent the existing players because they have revenue and the staff and support to do it, that that will change it in different ways.

Last point.  There is another downside to rules.  It focuses on abstract concepts like how fast you were going.

We don't really care about the speed directly.  We care about its implications for safety.  The standard focuses you directly on the safety issue, how do you drive safely.

And, in fact, I think one of the advantages of using standards instead of rules is it focuses you on the right things.  And unless you know there's a clear connection between the rules and the categories that you set up on those rules of what you're trying to achieve, it may cause you to focus on an abstract principle instead of helping consumers in the real world.

>> MODERATOR:  Thank you. 

So now it's Mike from the maybe camp. 

But before that, I want to make sure when you tweet, guys, can you also use the IGF 2015 hashtag with net neutrality. 

>> MICHAEL NELSON:  Thank you very much.  I'm glad to be here for several reasons.  I'm on the organizing committee for IGF and many years have been pushing for more debates.  So I'm glad Berin stepped up and organized this and very privileged to be on the panel with the others. 

I do apologize for leaving later and not being able to be part of the back and forth. 

The key issue for us at CloudFlare ‑‑ we're a start‑up.  We're one of the companies that has been trying to use the Internet to do something new.  We've been in existence for five years.  And what we do is protect 4 million websites from DDos attacks and provide those websites with strong encryption.

We do that by buying a huge amount of bandwidth all around the world and linking our data centers to hundreds of different telephone companies and cable companies and other providers.  So this issue is really important to us. 

The other thing that's really important to us is fighting against censorship.  And our company is committed to protecting people who have unpopular things to say. 

That's one of the reasons we exist. 

That's a human rights issue. 

I would argue that net neutrality at least net neutrality rules are really about economics.  It's about different companies trying to get advantage in the marketplace by using regulatory structures.

So I am very much in favor of net neutrality.  I think the Internet is so powerful because it grew up with net neutrality. 

What I'm opposed to is detailed rules to tell Internet companies, whether it's ISPs or CloudFlare or other companies, how to run their networks.  So that's why I'm a maybe. 

The other reason I'm a maybe is because my partner in the maybe camp, Bob Pepper, was my boss more than 15 years ago when I was at the FCC.  And he taught me just how complicated these issues are and why you have to realize the limits of regulation. 

So let me just provide a couple more pieces of ‑‑ other points that we care about. 

First, I think we tend to bundle together a lot of different issues under this banner of net neutrality. 

A lot of the focus is on the last mile which in some countries is controlled by one or two companies. 

But in the proposals we've seen put forward, including the one by my former employer, the Federal Communications Commission, there's also language about net neutrality for interconnection and for the intercontinental and transcontinental links.

That's a competitive market.

We've made deals with hundreds of different companies, thousands of different deals.  We have lots of choices and most of those are deals two engineers shaking hands, no lawyers in the room, because there aren't any rules for that part of the market yet. 

So of the two pieces, we're most concerned about that. 

The reason we can do these deals ‑‑ the reason we need to do these deals is because in each market we have different needs. 

This isn't a simple one‑dimensional agreement.  It's not just about speed and money.  There's all sorts of different complicated factors that go into these interconnection agreements, and net neutrality rules can't possibly cover all of that. 

What we see really works is competition.  As I indicated, in most markets we have vibrant competition in the interconnection market.  There's a few places like Australia and Taipei where we're paying five or ten times more than we should because there's not competition. 

But the market is working.  We have net neutrality now.  When you asked somebody advocating for net neutrality rules, it's hard to see an example that they can come up with. 

Net neutrality rules is a little bit like regulating morality.  There's 101 ways to be immoral. 

If you tried to write rules and tell people how not to act, they would always find another way to be immoral.

So let's not go down that path.  Let's instead use competition as the answer.  Let's open up the local markets.  Let's make sure that you have multiple ways into the home, have more spectrum, have new competitors who can get into the market.  In too many cities in America, the cities don't allow the roads to be ripped up more than once. 

Well, you just created a monopoly. 

And too many cities in America, apartment buildings provide only one provider for the people who live in the building so no competition. 

So we need competition.  That's going to be the answer. 

We need to follow the path that Jimmy Carter followed back in the '70s. 

Does anybody in the room remember CAB, the Civil Aeronautics Board?  This was a regulatory structure that regulated every aspect of airlines. 

They got rid of that under Jimmy Carter.  Suddenly, all sorts of innovation, airline prices decreased by more than 50 percent, and we got all sorts of new features.  More tiered pricing. 

In the old days, there was this idea that everybody is entitled to the same airline seat.  Not a good idea now and not a good market.

Same thing with trucking.  Carter also got rid of the Interstate Commerce Commission which regulated the cost of moving things. 

So let's look at how we can create competition, look at how we can realize the complexity of this market, and how fast it's changing, how fast companies like CloudFlare are coming into the marketplace, creating new services, new services that can't be anticipated by a regulator no matter how smart they are and no matter how well they can monitor the marketplace. 

Thank you very much.  And, again, I apologize for having to leave early. 

>> MODERATOR:  Thank you, Mike.

So now it's round two and we will start with Eduardo with the yes camp. 

>> EDUARDO BERTONI:  Thanks for the invitation and good morning to everybody. 

I would like to start with three reflections.  Three thoughts to frame my yes answer. 

The first question is something that we heard many of the workshop during this IGF which is what net neutrality is.  But this question is not always the same question of what net neutrality should be because some people start discussing net neutrality because they think what net neutrality should be, but what they are not discussing is what net neutrality is.  And when I'm saying what net neutrality is, I link it to regulations.

For some countries, net neutrality is what the law says it is. 

Maybe it's wrong, but we need to apply the law; and then net neutrality is what the regulation says. 

And, again, maybe we don't like what the regulation says because we think that net neutrality should be something different. 

And for the sake of example, when we at the Center for Studies of Freedom of Expression and Access to Information in Argentina, we provided testimony to the Argentine senate when they were developing the new telecommunication law; and we provided test testimony on net neutrality.  And we, at that time, when they asked us, well, how we should draft net neutrality in Argentina, we said, well, one way is to adopt or to take into account the model framework that is developed by the Dynamic Coalition at the level of the IGF.  This is one model. 

Again, they didn't follow strictly that model.  So that means that people do not agree on what net neutrality should be; and, at the end, legislators draft net neutrality the way they think it's better for the country.

And that might create problems because if you have net neutrality regulations in one country that are very different than net neutrality regulations in other countries that could create some sort of balance and problems globally. 

My second point is when we are discussing this issue, are we talking about net neutrality, or we are talking in truth about network management more than net neutrality? 

And I'm saying that because some years ago I read an article saying that today talking about net neutrality it's not good, it's not fair, because the truth is that the network needs some sort of discrimination. 

Some sort of management.

So it's better to discuss network management than net neutrality.  I didn't buy that argument.  But the truth is that in some way, the distribution of the package Internet are managed in different ways because of congestion, because of the different technical aspects.

So we need to think on that problem of network management when we draft net neutrality regulation that I agree that should be drafted.

And my final point for this short introduction, and then I'm going to make a couple of more points is why we are discussing ‑‑ and this is my point of view ‑‑ net neutrality in this IGF in almost all of the workshops that I enter.  Every time I enter a workshop there is a discussion about net neutrality plus zero‑rating policies. 

I think we are revisiting the net neutrality ideas and regulations just because we want to discuss these zero‑rating policies. 

And this is something that ‑‑ this is ‑‑ and I'm going to finish with this ‑‑ this is the situation in my region, in Latin America. 

Many countries start revisiting the net neutrality regulations because of zero rating policies.

Zero rating policies have been under discussion or implemented in Paraguay, Columbia, Panama, Brazil, Bolivia, Peru, Guatemala, and probably more countries. 

The problem when people start discussing their ‑‑ there are two main problems when people opposed to zero rating. 

The first one is zero rating confuse access to Internet with access to platforms.

So we don't like zero rating policies.  And my answer to that is, yes, it is a problem, but we can solve easily that problem if it is a matter of definition. 

Zero rating is not access to Internet.  It's not access to Internet and we fight for that.  That will be perfect, and we can solve that. 

But the second problem for zero rating is the net neutrality regulation that has been passed in the previous years.  And there are no, you know, equal laws in all the country. 

In my country of Argentina because of how the government or the senate passed the law on net neutrality, zero rating might not be possible.

In other countries, the situation is different and zero rating it not a problem with net neutrality regulation.

So my view is, yes, we need net neutrality regulations in advance, very well drafted, and thinking on how we're going to implement and control what we are saying in the regulations.

It's not easy to control discrimination in the distribution of information, but we need to do so in some way if we believe in net neutrality.  Thank you. 

>> MODERATOR:  Thank you, Eduardo.

Now, it's turn of Berin; and he's from the no camp. 

>> BERIN SZOKA:  Thanks.  Thanks, again, for having me. 

I want to recapitulate some of what Christopher said and add to it.  To start with, I want to get people to stop talking about the words net neutrality.  I don't think it's a particularly useful slogan.  I think it means as many things to people as there are people in this room. 

We can talk about human rights, but what actually matters is facts on the ground.  Human right to broadband is a great concept.  Doesn't mean anything unless you deliver people the services they want. 

I see two ironies in this debate.  First is the irony of people coming from the United States talking about an abstract concept that sounds great to everybody without really paying attention to the facts on the ground.

People from across the world have seen that in many contexts.

It's why people hate the International Monetary Fund, the World Bank.  Those are institutions that have pushed concepts that have been developed by academics in American universities without regard to what they actually mean in practice in the real world.  Here the same thing I see happening again. 

I see a second irony in this debate which is when we talk about the copyright, probably most people in this room would be offended by the idea of banning technologies in advance.  They say wait and see how things develops, be careful about meddling in what we don't understand. 

And yet many of those people have been convinced that for the sake of an abstract principle like net neutrality, that we do exactly that, that we should ban certain technologies and certain business models and in advance with ex ante rules.

I see that as a great irony. 

In both cases, what we ought to have is a more flexible approach that adapts to facts on the ground, exactly what Christopher proposed. 

So to recapitulate our position, it's that it's impossible to draw clean lines.  We can't put the Internet in a box.  We don't know how technologies or business models will develop and we should be more flexible in adapting circumstances as they evolve. 

The core of this concept is error costs.  The cost of getting it wrong here may be very high.

We can't predict what technology will be like in the future or, importantly, what business models might be needed to actually bring the Internet to people.  So zero rating is a perfect illustration of this. 

People talk about zero rating as if it's a conspiracy y Google and Facebook to solidify their control over the Internet; but, in reality, the little data that we actually have seems to suggest that it's the carriers in most cases who are pushing for zero rating plans.  And they're using it to compete with each other.

In other words, there's no market power here necessarily that we have on behalf of the big companies that are pushing for it.  It's usually carriers who are trying to deploy.  And usually the smallest carriers in the market who need nonneutral options to compete with well‑established incumbents. 

That was the point of the example that Christopher gave about T‑Mobile in the United States or MetroPCS which was the smallest carrier in the national market in the United States back in 2011 that was trying to serve poor consumers by offering them free unlimited ‑‑ those are offerings that serve consumers in ways that are very difficult for us to see in advance.

So let me emphasize the reality here from our perspective is drawing lines in a way that actually keeps up with how technology changes is impossible.  One. 

Two.  Regulators will inevitably be captured.

We see this all the time.  Regulation always begins with good intentions, but it is always captured by the companies that are involved, usually the incumbents who will find ways to steer concepts to sound great to defend themselves, and by political forces, people who actually have a vested interest in the status go. 

Our better approach, we suggest, is, first of all, how you regulate.  Don't write rules in advance unless you know there's a specific practice that is demonstrably almost always harmful.

Instead, rely on rules that are more flexible.  We would call standards, a standard of consumer welfare. 

But, also, importantly, where you can, rely not on a special regulator.  Rely on general regulation.  That is, if you can find laws that apply to all people in the market that are applied by a general purpose regulator, that will be less likely to be captured by special interests.  That generally means a consumer protection regulator, the antitrust, the competition authority.  Those are far more reliable than the special communications regulator which, by the way, have always regulated monopolies and are always tending to produce monopolies as Mike's example of the airline regulator in the United States.

And, finally, on that point, if you can, don't rely on the government to do this.  Use a bottom‑up process. 

There are processes that have been tried in the Netherlands and countries that have very strong commitments to net neutrality where the first resource is not to the government but to vehicles that dissolve disputes privately and then, only as necessary, go to the government. 

Finally, in closing, when you ask most people about net neutrality and they say yes, what they really mean is competition.  They want more choices in broadband.  They aren't happy with the status quo, and for good reason.  But they're different concepts. 

And we should ‑‑ as Mike was saying, we should be focused on that.  Even if we spent a small fraction of the effort wasted arguing about net neutrality or zero rating on deployment, we'd be in a much better condition today.

Why isn't there a Dynamic Coalition on deployment?  Why aren't we talking about the ways governments make it difficult to deploy around the world? 

Why is it that O3B, the satellite company that has their tent out there, they're providing wireless backhaul services connecting towers around the world with their satellite service.  Why is that that you have to use a connection from space to do that?  Well, the answer is because governments have made it so difficult to deploy good, reliable Internet access to connect towers and cell phones and all the Internet infrastructure of countries around the world.  Those are what we should be focused on.

>> MODERATOR:  Thank you. 

So after the introductions now is the fun time. 

Sorry.  I forgot the last maybe.  I was exited about the debate. 

So last, but not least, maybe camp.  Bob. 

>> BOB PEPPER:  Thank you. 

In the maybe camp, which is actually not about maybe.  I think it's ‑‑ you know, I clearly aligned with Mike.  It's the how. 

But that also is what we've heard from the yes and no.  So let me back up a little bit and try to pull up. 

What are we worried about?  What's the problem we're trying to solve? 

The problem we're trying to solve, and, again, when I was still at the FCC, we were faced with the problem where there was a carrier that was blocking SIP ports because they didn't want competition from Voice over IP. 

And we ‑‑ very quickly, it came up as a complaint in an informal complaint process.  We identified it.  We dealt with the operator that was doing that, the telco that was doing it.  It was a small, rural telco, and they stopped doing that.  And we didn't even go through the full process to impose a fine so we called it a voluntary contribution to the U.S. Treasury.

But what we learned from that was that there are problems that have to be solved.  There will be efforts for either ‑‑ in this case, it was both anticompetitive behavior and anticonsumer behavior.  So you have to worry about that. 

But the way we ‑‑ what we did, what evolved from that were a set of what we called net neutrality principles. 

And by the way, I agree with the phrase "net neutrality" is not very useful because it means so many things to different people; and Eduardo pointed out some of these, and I'm going to come back to one of those. 

So the fundamental principles there, again, they were not ex ante rules in advance they were a set of principles that then became over time embedded in FCC rule making.  But because of the idiosyncrasy, the uniqueness of U.S. law, three times the court said to the FCC, oh, you can't do it that way.

So what we ended up with in the U.S. because of the way the law is written ‑‑ it is not something that I would recommend anywhere else because I believe it is too restrictive in terms of ex ante ‑‑ but are principles. 

Number one ‑‑ and we did it from a bottom‑up consumer because it also protects competition.  Anybody should be able to connect any device to the network of their choice as long as it didn't harm the network. 

Two.  You have access and a right to have access to all content, assuming it's legal and that it's not, for example, child pornography. 

Three.  You have the right to download and use applications, because that also is about speaking. 

And number four, consumers need sufficient information to make informed decisions to be able to know what they're buying and then ‑‑ in terms of Internet access and then be able to see are they getting what they thought they bought.  Some basic principles.

And there were a couple of cases with complaints based on bad behavior in which the FCC responded and responded very, very quickly. 

But those things went to, again, what's the problem you're solving for, the concern about anticonsumer behavior and anticompetitive behavior.

And then there were some other things that evolved ‑‑ and Eduardo talked about one of them ‑‑ which is network management. 

That becomes important, and I'll explain that in a moment. 

And then the issue of innovation and flexibility. 

The Internet is moving very, very, very fast.  So you do worry about innovation. 

So what I like to do is go back to NETmundial, two years ago here, and in the looking forward document, there was a statement that said while there was no agreement on net neutrality, there was consensus and total agreement on what we referred to as the open Internet.  And the open Internet had a set of fundamental principles.  And, again, really important. 

The right to connect your own device.  The right to access any content.  The right to use any application.  The right to have information to make choices, but also recognize the importance of appropriate ‑‑ important word ‑‑ network management.  And, also, elsewhere in the document, there was a reflection and endorsement of innovation, and the basic principles within NETmundial, principles ‑‑ not the going forward document, but principles was probably the strongest statement about the freedom of expression in any similar document because it talked about the right to speak not just on the Internet but in any medium. 

So if you incorporate those principles, they are principles and use techniques that can address bad behavior to protect consumers, competition, innovation, freedom of expression.  I believe that's a much more effective way of doing it than having ex ante rules that during the discussion session, I will give you concrete examples of how ex ante rules are harming innovation and inadvertently harming consumers, particularly because we don't know and we can't predict how the Internet is going to evolve. 

>> MODERATOR:  Thank you, Bob.  So now it's the fun time. 

So when we first started, I told you this is different from the other net neutrality panels; and I wasn't lying because we have ‑‑ this is going to be a debate.  And we have like six panelists in three camps; and now they will discuss, and they will ask each other questions. 

So each group has eight minutes to ask questions and answer.  So I just start with Mike and from the maybe group. 

>> MICHAEL NELSON:  Thank you.  I do appreciate that since I have to leave in five minutes for another panel.

One of the things that I like to compare this debate to is airlines.  I mentioned the demise of the Civil Aeronautics Board because there wasn't enough competition or choice. 

So for both sides, I'd like to ask what lessons you draw and in particular should there be airline seat neutrality so every seat sells for the same price and they're all the same class of seating?  That's a facetious question but ‑‑

>> HAROLD FELD:  I will begin with my own observation of irony at having been invited to attend the panel, been reprimanded for coming to express my opinion on the subject, and to find instead what is offered as being the appropriate path to embrace the now discredited Washington solution of the 1990s to simply deregulate our way to deployment and freedom. 

So we all have our little ironies that we appreciate. 

With regard to the specific question, I have to say that having spent some considerable time in airlines at the moment, having been subjected to their nondisclosure to what has become the eternal slippery slope, especially for a gentleman as portly as myself with regard to seat size, I have to say that I do think that some basic minimum standard enforceable so that I as a consumer can understand what I am going to buy, can make reliable purchases is, in fact, pro consumer and necessary. 

I think the airlines provide an incredibly instructive example, particularly since we have reconsolidated our way back to a noncompetitive state which shows the complexity of the problem, the need for some regulatory oversight, some basic ex ante rules to provide sufficient certainty while balanced against too rigid a regime. 

I am criticized apparently for ignoring facts on the ground, but I would say that the realities on the ground reflect the complications of reality and show that the rules in net neutrality, at least those adopted in the United States as well as those adopted in other countries, arose in response to specific circumstances where they are subject to fine tuning and discussion. 

I do not see that the Internet has crashed and burned because the engineers are no longer capable of concluding deals.  But providing some basic rules about what may not be included in the handshake deals or requiring that you talk to people you might ‑‑ your bosses, who are not engineers, might tell you not to deal with, serves to keep the Internet running smoothly. 

>> BERIN SZOKA:  So this is Mike's use as he sees fit.

>> MICHAEL NELSON:  No.  I asked the question.

>> BERIN SZOKA:  Indeed, but you should stop me when you decide I haven't responded.  It's just as Harold who didn't respond to your question. 

But, look, since Harold brings up the example of ‑‑

>> MICHAEL NELSON:  It's really a good thing you're over here and he's over there. 

(Laughter)

>> BERIN SZOKA:  Oh, that was quite deliberate. 

Since Harold brought up the example of airlines ‑‑

>> HAROLD FELD:  I didn't bring up the example of airlines.  That was the question. 

>> BERIN SZOKA:  Okay.  Fine.  Mike asked that question, and Harold responded. 

The reality is that airline prices have fallen 50 percent in the last 30 years.  That's just a fact.  The fact is that changing how we regulate airlines brought them within the reach of people who could never afford to fly. 

So it's not that they were deregulated.  The government of course still regulates them, still deals with safety issues and consumer protection.  It simply changed how it regulated them. 

And what it did was to abolish the special regulator that had made airlines a monopoly or an oligopoly.  And, instead, we have a competitive market.  There were no new airlines started while we had a special purpose regulator.  Today, we have a different approach. 

To apply that to the example of the Internet, I'm not saying that there isn't a role for government.  Harold puts up a straw man when he says that government ‑‑ that we're claiming that there shouldn't be any role or arguing for complete deregulation.  We're simply saying be careful how you design rules.

Let me give you examples of rules that I think we should have ex ante. 

Don't lie to your customer. 

Good rule to have because we know that lying to your customer is really not ever a good thing. 

But when you get beyond that and you get into the more difficult tradeoffs about how you regulate networks and especially areas where we don't know what the effects are and the tradeoffs about zero rating and whether it's important for a carrier to be able to offer a nonneutral offering to attract consumers, there we need flexibility.  And that's where we want a standard that we don't write in advance.  We apply it as it goes. 

>> CHRISTOPHER YOO:  So if I can try to take this back to the Internet, but drawing on airports and airplanes, a classic example is something almost everyone experiences, and it's not a U.S. example. 

When you go to the airport, sometimes you get free WiFi.  Sometimes they make you watch an ad for a particular company that's supporting the WiFi.  Or you have to pay for the service in order to get on with the right to see it without having an advertisement.

It this discriminatory?  Is this favoring one provider over another, one advertising point? 

Well, you can say anyone can buy that ad and make it open in that way from a competitive standpoint, or we could regulate and say you cannot have this particular practice. 

I actually don't know, because the airports are doing different things, what ultimately will be the best solution.  And maybe different ones will try different things. 

But what's striking to me is that I think many consumers find this is an acceptable compromise and they would rather save the money.  It increases their options because they have the ability to do that. 

The bigger concern is this. 

If we have rules, we focus on this practice and say is it discriminatory?  To me the real question is is this practice allowing this practice good for consumers. 

Does it increase their choice? 

Does it allow them to reduce the amounts they pay because if services don't count against your data cap, you can reduce the amount you pay.  And what is the service that's buying in?  Do they pay money or it is an open platform?  T‑Mobile makes it open to any music service.  It's just something they're trying to differentiate it.  They're not trying to get money out of it. 

And in one of the cases in Slovenia, it was a small cloud service, not one of the big giants, called Hangar Mapa that was trying to push its way in and trying to find a way to get a purchase in the market that's crowded by established players. 

So by looking at was it good for Slovenian consumers, not is it violating some abstract concept of discrimination, to bar zero rating of Hangar Mapa cloud services? 

I think focusing on the consumers and the real issues of how this plays out will make a much better policy.  And I think ex ante rules focus you on the abstract principle and not on the real thing.

>> BOB PEPPER:  I have a little bit of time to ask more questions.  Picking up on ‑‑

>> MODERATOR:  Before that, Eduardo wants to ‑‑

>> BOB PEPPER:  Oh, okay.  I'm sort of trying to pick up on Mike because he had to leave.  I think we have a few more minutes left of maybe to ask questions of yes and no. 

But I want ‑‑ maybe Eduardo. 

So one of the things that Harold said was that, you know, we need ex ante but also need flexibility.

Christopher just said you do need some regulations.

Berin said we absolutely have the ‑‑ on an ex ante ‑‑ you can't lie to people.

So for both sides, what would be ‑‑ for the no, what would be appropriate ex ante?  And for the yes, how do you ensure flexibility? 

>> EDUARDO BERTONI:  Thank you for the question because it's exactly what I want to raise right now.  I think that and I will ask another question. 

To answer your question, for the yes is in a regulation, you can include basic principles, in fact, basic human rights principles, that are in international standards, in constitutions like the nondiscriminatory principle, noncensorship, freedom of expression, so on and so forth. 

So some of the regulations of net neutrality include these kind of principles.  So one of my questions is why we are going to be against a principle, net neutrality, that translate into this area some of the general principles that defends basic human rights like the nondiscriminatory.

I think that at some point, and I'm going back to my first words when I talk here, is my feeling is that people who are in the camp of the no are more in the no of the regulation of what I call the network management. 

And this is different.  I would agree with that. 

I would agree that if we want to have in advance a very specific regulation on how to manage congestion, for example, maybe that will be a problem. 

But if at the end of the day, the way that you're managing the network is against one of the basic principles, well, you are in trouble there. 

But I would say that if you ‑‑ if you are in the no camp because you believe that the management could not be regulated in advance by the law, the management of the traffic, I would say, well, I'm more on a maybe.  But I think that it is possible to put a regulation that says yes with basic principles that, in fact, are principles that we recognize for any content that we want to deliver. 

>> BOB PEPPER:  Sure.  So what would be ‑‑ thank you, Eduardo. 

>> CHRISTOPHER YOO:  Easy ones that are established.  Transparency, making clear what the principles are, making people honor them, basic consumer protection terms.  Fraud is always bad.

There is room for baseline technical capabilities to make sure things operate as expected.

We need to make sure that those standards morph over time. 

What we're discovering, one of the things that's badly misunderstood is if we want an open environment with third parties providing services, you have to say ‑‑ if someone brings an app and says I'm doing things that comply with this system, you have to have a way to validate they're telling the truth, because a lot have incentives to cut corners.

There's a mapping function.  Engineers recognize that geosatallite mapping data is more accurate than the other way which is you look at the local cell towers, measure distance, and figure where you are.  And they use the data on an ongoing basis to improve the information by recursively feeding it in, but they require you to code how you got the data, from a satellite or from cell tower.

There was one company that said their product is to good that even though we do it from cell towers, we're coding as satellite data because it's the quality.  That's bad for consumers.  You have to have a means and the architecture for making sure that people are doing what they say the architecture requires them to do. 

If we're going to have a world in which it's not provided by all one company, you have to have the means to validate and testing to make sure the system all works properly. 

>> BERIN SZOKA:  I'm dying to hear what Harold says about building flexibility into his approach.  But let me just tell you how I'd do it, because in addition to what Chris just said, I would say two things. 

Number one.  I really don't know what the right answer is, and that is fundamentally my point.  But I'll give you two suggestions about how to do it. 

Number one, in many cases when we talk about regulation, it's not necessarily about banning it.  So the question might be you could have rebuttable presumption.  You could say we think this practice is likely to be harmful.  We're not sure so we're not going to ban it completely.  But we will at least set a default presumption that it might be harmful and allow it to be rebutted. 

That would be in general the most I would want to do with business practices that are not deceptive or clearly, clearly harmful.

Number two.  Because I don't know what the right answer is and because it might vary from country to country and context to context, that's why I emphasize my point about who decides. 

And I would like to see these decisions made not generally by the legislature or a regulator, whether special regulator or general regulator, I'd like to see these decisions made in a multistakeholder context whenever possible so you have a standard that can be enforced.  It would make the regulator's job much simpler because then consumer protection regulator enforces the standard, holds companies to it; but the details which are hard are worked out by the multistakeholder approach with people like Harold sitting at the table. 

Harold, what do you think? 

>> HAROLD FELD:  I'm not sure why from Berin's perspective I'm the only person at the table, especially since I basically agree with Eduardo. 

I think we have history of regulation being done right.  We have history of regulation being done wrong.

We have history of where the creation of flexibility, the problems of postdoc regimes typically go to ‑‑ having argued this previously in this very context go to complaints about uncertainty, complaints that without some specific clear rules about what is and isn't permissible, that companies do not know how to invest, the problem of without some clear guidelines, every complaint begins with a fight over whether the conduct ‑‑ whether the agency even has reason to engage in an examination of the conduct. 

You know, these are problems where at least with regard to everybody but Berin, we are talking about degree. 

And I think that we can argue where in the swing of the regulatory pendulum things ought to go.  But Eduardo is quite correct. 

I will just add one response to Chris which is that I think determining consumer welfare is not so simple a matter. 

It is not simply a microeconomic what lowers a cost.  And from a pro competitive standpoint, it is not simply can you determine whether there are some set of balancing efficiencies.  This is why I believe that certain principles like the no blocking principle must be absolute, and that where we invite balancing for a very limited class of principles such as either deceptive practices or blocking specifically, that those practices are inherently harmful.  And I think we do have sufficient information to label a very small class of conduct like the blocking of VoIP calls and like the blocking of any phone call to be inherently harmful. 

>> MODERATOR:  Just a second.

So before that ‑‑

(Discussion off microphone)

>> MODERATOR:  Just to clarify, so we have half an hour left, and I want to leave 15 minutes to each group to make their conclusions.  I'll skip my questions because this debate is going really well.  So you have 15 minutes.  And I want to divide it to three, like each group have five minutes to ask their questions and answer the other's question. 

So I'm just like resetting my stopwatch.  And your five minutes start.

>> EDUARDO BERTONI:  But I would be less than five minutes.

>> MODERATOR:  Use the time to ask questions.

>> EDUARDO BERTONI:  I think we're almost convincing Berin to come and sit on this side of the desk because what I just heard is that you do not opposed to a net neutrality regulation.  You oppose more on process.  You don't like regulation coming from legislation.  You don't like regulation coming from regulators, specific regulators or general regulators.

This is what you said, and this is what I understood.  And you prefer a multistakeholder mechanism that implement the net neutrality principles.  Maybe I understood it incorrectly, but this is what I heard. 

But if that is the case, we need to think in advance on how you are going ‑‑ or how we can implement that idea, the idea of multistakeholder involvement in the definition of net neutrality in advance. 

And I would say that in a democracy, we have that mechanism.  In a democracy, that mechanism is in congress. 

I totally oppose i the regulator or the legislature start drafting network neutrality legislation without hearing all the stakeholders involved. 

But as soon as they invite everybody, the business sector, civil society, the human rights defenders, so on and so forth, then we have to have something coming ‑‑ an output coming from that process, which could be a multistakeholder or should be or have to be a multistakeholder process.

So I think, Berin, that you can sit on this side and come with us and say, yes, I want a regulation, but I want regulations coming from a specific process. 

Thank you. 

>> MODERATOR:  I'll ask the no camp to ask their questions and answer. 

>> BERIN SZOKA:  Well, first of all, thank you, Eduardo.  It's gracious of you.  And I'm glad we're not arguing and this is a friendly debate.

I want to share with you just to explain my answer. 

I'm going to read you something, and then I'll tell you who said it. 

Open transparent multistakeholder processes, when appropriately structured, can provide the flexibility, speed, and decentralization necessary to address Internet policy challenges. 

The document goes on to say flexibility is critical to resolve Internet policy issues and talks about how technology changes.

Now, you might think that was coming from TechFreedom, but that is actually coming from the Obama Administration's 2012 consumer privacy bill of rights white paper.

There's a lot of good stuff in there about ‑‑ talking about how difficult it is for Eduardo's preferred model of a congress to come together in a democratic process and write rules because this is not the 19th century.

We're talking about technology that changes and issues that are difficult for the political process to get right. 

So the reason that I prefer these issues to be resolved not through government but whenever possible through multistakeholder processes, and sometimes through industry codes of conduct, and it depends how that works out, is because those alone, as the Obama Administration said, provide flexibility.

So to be concrete, in the United States, this debate about net neutrality has been insanely frustrating for me because it's been a unnecessary debate to a large degree. 

There is no company in America that I'm aware of that actually has said they intend to block user traffic. 

In fact, they've said consistently broadband providers have always been willing to say of course we agree to basic transparency disclosure, no blocking requirements. 

The question is always on the margin. 

What does it mean to block?  How does it apply in more difficult circumstances? 

Those are the areas where you need flexibility, which is why, Eduardo, when you hear me agree with you ‑‑ and in some sense I do, because to a large degree, these debates could be avoided and in many countries like the Netherlands have been where you've got a multistakeholder code that is agreed upon and then enforced by the government where the government doesn't have to do the hard work of coming up with strict, rigid, ex ante rules to deal with the cases.

So does that make sense, my answer? 

>> PANELIST:  So what's your question for us? 

>> MODERATOR:  No, no.  I want audience to be involved in this debate.  Hold on.

I want audience ‑‑ no.  Just type your questions.  We are at IGF.  Use Twitter.  We have hashtags.  Like if you have questions for the panelists because they still have like ten minutes. 

>> BERIN SZOKA:  It's IGF 2015 and net neutrality.  If you tweet on both of them, we're watching on both of those hashtags. 

(Discussion off microphone)

>> BERIN SZOKA:  Well, sure, but we've been using both hashtags; and there's not a lot of discussion if you use both.

>> MODERATOR:  Yes, we are watching them.

So if you have questions for the panelists, I'm going to pick two interesting questions; and I want you to be involved. 

But in the meantime ‑‑ so you have time as from the member of maybe group.  If you have questions for the panelists or if you want to ‑‑

>> BOB PEPPER:  I was asking questions before as the maybe group for both sides.  I just wanted to maybe other than ask questions, I had comments.  But can I hold those if you want.

>> MODERATOR:  Go ahead. 

>> BOB PEPPER:  So let me give a very concrete ‑‑ because I think Eduardo said, well, maybe a little bit in the maybe camp because of sort of the technical issues and the network management issues. 

Again, I think part of the problem is that the phrase "net neutrality" has been so emotional.  That's why I liked what NETmundial did in talking about the open Internet and open Internet principles.  If we can turn down the heat a little bit and talk about what it is we really want in a positive and also what we don't want in the negative, I think we can actually come to a much closer consensus.

But let me give you a concrete example of how things move quickly. 

Eduardo talked about nondiscrimination as being important and when we're talking about nondiscrimination when comes to people, yes, that's sort of tied back into human rights. 

When we're talking about making sure that anybody has a right to speak, freedom of expression, content that can go end to end without being blocked, all of that. 

But the word "nondiscrimination" in telecom law ‑‑ and by the way, in the United States, it's not nondiscrimination.  It says carriers may not unreasonably discriminate, because there's a recognition that there are different types of traffic that need different types of management and to be approached differently. 

So, you know, we do this forecast every year which is about everything to do with the Internet, the number of people connected and the traffic and what it looks like.  And so this year ‑‑ and we've been doing that for ten years.  It was a five‑year rolling forecast, and we actually report back how did we do on the forecast.  And it was close.  So I'm confident in numbers.

So I think we're at a point of a tale of two networks, and it's not just about congestion. 

We know that today video is the predominant form of traffic.  And we're forecasting in 2019 on fixed networks globally, it will be over 80 to 85 percent of traffic.  Even on mobile, on cell phones, it's going to be 75 percent of traffic.  So it's all about video, video, video. 

It's going to be download video, streaming video.  It's going to be two‑way, Skype video.  It's going to be you creating video as well as consuming video. 

So it's producing video as well as consuming.

But the other ‑‑ the second tale of the two networks is the ‑‑ if you look at the things being connected, the fastest‑growing category globally are machine‑to‑machine devices.

So, globally, we're forecasting 43 percent by 2019 all of the things connected to the net are going to be machine to machine. 

And, in fact, in some countries ‑‑ like Japan will be 68 percent, U.S. will be 58, and Korea 72 percent.

What's interesting is although it will be 43 percent of the devices, it's only 3 percent of the traffic.

So we're looking at a completely different type of set of things being connected with different sets of requirements. 

Some are going to ‑‑ most of them, by the way ‑‑ some are going to be continuous, like the security camera, right, when we came in.  That's continuous.  It's broadband.  It has power.  It has Ethernet. 

But most of the things are going to be little things that are going to wake up and chirp little bits of data, and go to sleep.  Repeat.

Some are going to communicate over short distance, some long.

Some broadband, some narrow, some continuous, some chirp.  Some are not going to be sensitive to latency.  Others, like the device connected to the air bag in your car, requires very low latency. 

So the set of requirements for all of ‑‑ by the way, we're forecasting 2019 there will be 10.5 billion of the M‑to‑M devices connected.  So it's about complexity. 

So you need to manage for both consumption and for both congestion.  That's the traditional thing we think about.  But now we're beginning to see this whole new class of things that are going to have to be managed for complexity as well. 

And it's important that we have the flexibility in whatever the rules are to be able to do that. 

And what we don't need are well intentioned ex ante rules that limit flexibility for the innovation and the kind of networks that we need in the future while at the same time preserving the principles of the open Internet.  It's a false choice to say that it has to be either or.

>> MODERATOR:  You have 30 seconds.

>> CHRISTOPHER YOO:  So it's funny.  I realized something that I like things about debates.  They can also cause conflict.  They can also narrow and point out where you agree. 

I actually realized a little while ago I wrote the second article on network neutrality in 2004. 

Tim Mo coined the phrase in 2003, "They invited me to write a response."  He wrote a response to me. 

So it's partly my fault to get the debate started.  It's not my fault for keeping it going as long as has.  I will not take that responsibility.

BERIN SZOKA:  So we blame you. 

CHRISTOPHER YOO:  Well, I'll take some responsibility.

And in that case, I endorsed a version of network neutrality regulation but based on ex post case by case.  And I think in general, most companies will acknowledge they committed to the open Internet and willing to accept a certain amount of regulation as long as it's well formed. 

Harold brought up a question which is no blocking.  That's the rule.  I was tempted to say that's one of my rules that I would accept.  Then I started thinking of technical problems. 

There is a problem called route flapping.  If a route is available and then it goes down, and then it comes up and then it goes down, it causes problems with updating the routing tables.  So ISPs will take down that site. 

Are they blocking a website?  Yes. 

Why?  Because it's creating technical problems for the network that's causing enormous problem for updates. 

Okay.  So I'm realizing we can make a technical exception. 

They block known sources of malware and threats into the network.  And they'll block sources of DDos and attempt to mitigate different aspects. 

And then the last thing that's really interesting to me is the case that Harold brings up which is VoIP, I'm really tempted to do as well; but one of the technical realities is that because IP is a general service and voice is a specific service, voice is 15 to 20 times more efficient in bandwidth than VoIP. 

So one of the technical realities in a very low‑bandwidth network, that's a problem. 

Now, for SMS response?  Probably not.

I mean we're not talking about huge amounts of bandwidth.  But would I say no blocking under any circumstances?  I don't know that I can say that. 

>> MODERATOR:  You can ask your question.

>> HAROLD FELD:  I will have a quick response because I know I want to get to the other questions. 

But my very quick response is I see a great consensus and it is a question of whom do you fear more and whom do you trust more to handle the complexity.  My experience ‑‑ my perception from the no is we agree in principle, but we believe government is inherently corrupt, it will be inherently captured, and inherently incapable of handling complexity and therefore should be profoundly limited in scope, and the rational order of these things should be that it be handled primarily in the private sector.

Then through multistakeholder processes and only in implementing the multistakeholder processes or as a truly last resort should government ex ante regulation be provided.  I believe the world is somewhat more complex than that.  I believe that government and industry and civil society interact in a variety of different ways. 

I have seen the black swan which is you tell me nobody will block.  This is the problem of a black swan.  How many swans do you have to see to prove that swans are not white?  Answer, one black swan. 

Comcast blocked.  There's no getting around it.  Wireless carriers traditionally blocked and many still reserve the right to do so and on occasion do so. 

So the argument that has been put forward that it will never happen or only happen in the cases of network security are demonstrably false because they have occurred before.  We can argue about the likelihood of again, but they've occurred.  So they're not impossible.

The issue then becomes where is the arbitrator of this.  As somebody who sits in civil society, I will say, yes, government is frequently flawed.  Government is frequently captured.  There are great asymmetries of the ability of citizens and others to argue with large corporate powers in the context of government regulation; but, ultimately, the only way that we can have an effective means by which these principles are enforced is through government regulation.  And, therefore, on the whole, I do prefer specific minimum ex ante rules that set the reasonable expectations.  And then we move to the complex cases that you have raised which are no more complex in this context, I would argue, than in numerous other contexts where we have resolved these complexities. 

All right.  My question for you is:  So what in the basic regulation ‑‑ never mind Title II.  But what in the basic regulation that the FCC adopted, preferably the 2010 regulation and the 2015 regulation, which for those not familiar with the general context, the 2010 was a no harmful discrimination standard.  The 2015 has three bright line rules ‑‑ no blocking, no degradation, no paid prioritization ‑‑ and then a fourth catchall, no getting around these three rules through other discriminatory means.

So what about those do you think are bad, that you see them as inappropriate? 

>> CHRISTOPHER YOO:  What I would say, it's hard to say in the rules themselves because they punt so many decisions down the road.  The devil is in the detail. 

So one word that's not been spoken in any of the net neutrality panels I've seen here yet is specialized services. 

This is the entire focus of the EU dispute between the European Parliament and the European Council.  And the scope of that is going to determine a great deal about how much flexibility is left in the system.  So I'll give you several examples of common things that you use that have specialized services.

If you have a proprietary video system on DSL in your home, they use specialized service.  Otherwise, the only choice is to lay fiber; and we know that's very expensive. 

If you have view IP service that meets ITU standards of quality, which is a 300 millisecond delay, you are using specialized services.  You're using dedicated bandwidth to do that. 

We've been trying to do VoLTE, Voice over LTE over fourth generation phones.  If you use VoLTE, they get great quality.  They can only do it by reserving bandwidth.

And one the great mysteries in all of this if we regulate this in a way that permits ‑‑ and I'm sure there are other things out there we haven't thought about.  Those are the everyday things that most of us use constantly to do this. 

If we regulate in a way that permits new experiments like that to occur, I think that we can work within the frameworks of the rules because of things like specialized services and reasonable network management.

What the question is is what will happen ‑‑ if those are regulated in a restrictive manner, we're not going to be able to do that. 

And when you ask who do you trust more, I think the Internet economy has thrived because of decentralized innovation in ways that no one dreamed possible.

We all invested in Facebook in the beginning because we knew it was going to be huge.

We didn't ‑‑ and I think we're all smart.  No one is so smart to know what's going to succeed in advance. 

So what we have to do is create a world in which things are flexible. 

I want to throw in one last point.  We also don't talk much about government as legitimate stakeholder.  And particularly in network neutrality debates, they are often put to the side. 

And I will say to Harold there are a bunch of interesting issues.  One thing that's important to governments is taxation.  You see all these people and companies taking their operations off shore and wiping out their tax base.  There are in international law enforcement some legitimate concerns that need to be taken into account in this debate which are rarely discussed, and I think that's a problem. 

>> BERIN SZOKA:  If you remember two words from what I say today, remember these:  flexibility and discretion. 

A point made by Christopher is even when you try to write rigid rules in advance, there will inevitably need to be some room for flexibility.  But what it actually ends up meaning if you try to write a rigid rule, is you end up ‑‑ the regulator ‑‑ claiming discretion in the future to decide.  And it becomes arbitrary.

The regulator says, well, we know doing this will break important functions so we'll leave ourselves discretion in the future to decide to how to regulate things like special services.  That is inevitable. 

Whether it's reasonableness for blocking or for the example Christopher mentioned or specialized services, at the end of the day, the regulator will decide probably arbitrarily in that model how to make sure that their approach to regulation doesn't break things for consumers. 

My fundamental point is that it not a good model.  I do not trust a regulator, especially a special purpose traditional telecommunications regulator that's been designed for monopolies to have that much discretion. 

That's why I prefer to deal with the problem of discretion in several ways.

One, as I keep saying, let's preferably have the regulator not be a special one but a general purpose regulator that is applying standards of general applicability.  And the standards of general applicability that I would prefer are those that come from consumer protection and antitrust law.  Because then when the regulator makes decisions, as it will, it will have to have discretion, the decisions will be less arbitrary. 

The regulator will be able to point to standards that are developed that, for example, tell us when a practice is so unfair that it should be banned to consumers. 

We have a test in the United States for that.  We say does the harm outweigh benefit to consumers?  And is the harm something that consumers themselves can't reasonably avoid.

If you do cost benefit and if the harm is something consumers can't avoid, yes, the government can declare the practice harmful.  But that's important because that standard of consumer welfare constrains the regulator's direction. 

So that's really, at the end of day, what I'm saying we need.  And that's why it's important that the regulator ideally have its constrained, and that, where possible, you don't send these decisions to the government and you have them made in a multistakeholder process that both crafts a more flexible code and also handles the arbitration on a case‑by‑case basis. 

>> MODERATOR:  Okay.  So you guys will have one minute each for your concluding points.  And I think we will run five minutes late so that's your last chance, guys.  So if you want to ask a question to the panel, please tweet it now or it will be never.  Or you can you find them during lunchtime. 

So you have like 30 seconds to ‑‑

>> BOB PEPPER:  So 30 seconds to answer Harold's question. 

I actually ‑‑ again, this should not be about U.S. and U.S. law. 

I agree with the 2010 principles that Harold mentioned ‑‑ that's why maybe I'm in the maybe ‑‑ because effectively they were the open Internet principles.  What happened, again, as I mentioned earlier, because of the uniqueness of U.S. law, the FCC was told you can't do it the way they did it.  And they ended up doing it a different way.  The different way that they did it, which was traditional telecom law, which has all kinds of constraints ‑‑ Harold said let's not talk about Title II, but that's what that is ‑‑ that's where I don't agree. 

It was not the principles.  I agree with the principles.  They're important.  But it's the way in which it's done also is extremely important so that it maintains the flexibility.  But this open Internet principles broadly including appropriate network management, innovation, protecting consumers, competent, sufficient information to make decisions and be able to connect your own devices so you can go anywhere on the Internet, those are really important because those support these other broader principles, for example, of freedom of expression and human rights, et cetera. 

Last point briefly.  You can't have really binding enforcement without government. 

So there needs to be a governmental mechanism even if decisions are made within a multistakeholder process.  If somebody violates the principles and it's proven, then the question is what's the enforcement, it could be by contract; but, ultimately, contracts are going to be enforced by government. 

So it doesn't mean there's no role for government.  It's what the appropriate role to maintain the flexibility.

>> MODERATOR:  Okay.  So there is only one question coming from the audience.  I'm going to ask that question to you guys.  And you guys have like 30 seconds each or like in total like one minute to answer the question.  Whoever wants to answer, can do so.

So the question is:  Are there examples of network providers restricting freedom of expression justifying ex ante regulation? 

Harold? 

>> HAROLD FELD:  Well, Comcast, the largest provider in the United States, explicitly blocked peer‑to‑peer applications and particularly BitTorrent for a period of I believe it was a year and a half.  They ultimately submitted a detailed explanation of their conduct including the manner in which they blocked it which simulated a man‑in‑the‑middle attack. 

This is on file at the Federal Communications Commission.  It is available online.  So we do not need to guess.  Comcast signed it under the penalties of perjury, so we know it is true.

That would be one significant example. 

There was evidence that other providers at the time were implementing similar blocking schemes and using deep packet inspection in order to be able to block traffic.  But they ceased when the Comcast complaint began. 

Other examples, see me in hall. 

>> BOB PEPPER:  So what's interesting about the example that Harold uses, which is absolutely correct, is it proves the point you don't need ex ante regulation.  Why? 

Because the FCC had its principles.  As soon as it became evident that Comcast was doing it, it was stopped by the FCC.  Stopped dead in its tracks.  There was no ex ante regulation. 

It was the principles, and it was the application of that.  And the FCC intervened immediately and stopped the bad behavior. 

So, again, it's not that there won't be bad behavior.  We know there will be.  It's how do you do it.  And I think that that's an example.  But it's also the example of why you don't need ex ante regulation if you have good visibility and, by the way, nothing is hidden on the Internet. 

>> CHRISTOPHER YOO:  So I have to correct ‑‑ first, what Comcast did was bad.  Don't take what I'm saying to contradict that, because they didn't represent it correctly. 

Actually, it's not correct to say they blocked.  They throttled.  They used TCP resets, and the details do matter. 

What I would say, though, the other examples that are commonly given actually are often not Internet examples.  And I'm happy to talk about them.

The other thing is ‑‑ the question was:  Is this an example of them doing it against free expression?  You can say they were attempting to curb speech and it might have had that effect.  It was really a business decision.

You could say they were protecting video because they're a video provider. 

The other reality is BitTorrent is engineered in a way that it expands to fill all available bandwidth in ways that machine‑to‑machine things do.  And it was ‑‑ well, they can.  Not all of them do. 

It's a problem because without incentive, they had no means of metering BitTorrent.  And the problem is if you're in a node with 50 people, five BitTorrent users can clog your entire site.

So there's a couple of explanations here.  It's complicated to get to the bottom of them.

There are examples that I think are problematic.

There is a reported example in Canada where a network provider blocked access to a labor union site with whom they were having a labor dispute.  I can't deny these are possibilities.  An interesting question is what's the right way to do it, and is it so prevalent that we'd create a general network neutrality rule as opposed to one in certain circumstances.

>> BERIN SZOKA:  I know we said we'd take questions from the audience but ‑‑ is this working? 

I would just briefly say that the other thing the Comcast example proves is there were other laws on the books that could have addressed that problem.  Our consumer protection regulator could have addressed that case.  What Comcast was doing was clearly deceptive.  It was a purely political decision that the person who was running the FCC at the time decided he was going to decide the question rather than allowing it to be decided by the general consumer protection regulator.  I think it was a mistake, and it was the thing that fundamentally derailed the conversation in the United States which could otherwise have been about how we use existing laws.

>> MODERATOR:  Okay. 

So we have three very interesting questions from the audience.  One of them is for you, Harold, specifically for you.  So you are the most popular guy on this panel. 

So first one is like it's for all the panelists.

Network neutrality is just a glorious indefinite term supported by everyone, whatever they define it to be.  What is your definition? 

And you have only 140 characters because you are answering the tweet question so you are assuming that you are answering the question on Twitter. 

Okay.  Very short.  Who wants to start?  What is your definition of net neutrality?  Bob? 

>> BOB PEPPER:  I've already said it.  It's the open Internet as described in the going forward document from the NETmundial. 

>> MODERATOR:  Great.  Eduardo?  Harold? 

>> EDUARDO BERTONI:   When my students ask that question to me, I say ‑‑ my first reaction is, well, it's not an agreement on that.  But I always read the model framework designed here at the IGF level. 

>> CHRISTOPHER YOO:  Business practices that disadvantage certain actors in ways that hurt consumers.  That means not justified by cost and not justified in a market where it's a good business strategy to differentiate what you're offering instead of doing the same thing as everyone else. 

>> BERIN SZOKA:  That's how it should be defined; but, of course, we're the no camp so we're saying no to the prevailing definition, which is very different from that.  The prevailing definition essentially, if I put it simply, is we know what the Internet should look like.  It looks like something that we think of today that should never change, and we should write that into law today.  That's why I'm saying no. 

>> MODERATOR:  Okay.  Harold, this is your question.

>> HAROLD FELD:  Didn't I get to answer the other one first? 

>> MODERATOR:  Sorry.

>> HAROLD FELD:  I was just going to say I think John Oliver speaks for me. 

>> MODERATOR:  So your special question is coming. 

By whom ‑‑ at which level would you like to see regulation?  Nationally?  Regionally?  International? 

>> HAROLD FELD:  I believe that as an initial matter, it requires national legislation.  I think every country is different. 

And the questions of ‑‑ within the scope of the general principles that we are discussing how they're implemented is something that will need to be decided on a national basis. 

I do think that there appears to be broad international consensus, and I would hope that things like multistakeholder processes which we certainly support and applaud inform the process of national legislation. 

>> MODERATOR:  Okay. 

The last question is for everyone. 

Does no practicing standouts can be enforced without government involvement? 

>> PANELIST:  Can you say that again.

>> BERIN SZOKA:  If I may, I'll answer.  The question was:  Doesn't government need to enforce standards? 

Of course.  That's what I've been many times.  Please don't misunderstand me.  I'm not saying government shouldn't have a role.  I'm saying it really matters what that role is.  And my best case scenario is that government doesn't write the standard, and a telecom regulator is not involved; but you have the general purpose competition and consumer protection authority that enforces standards. 

That is the best case scenario for me.  But, of course, that means there's still a role for government. 

>> MODERATOR:  Okay.  Christopher.

>> CHRISTOPHER YOO:  Because this the IGF, I would say, though, they're not the only actor.

Industry best practices can actually resolve a great deal of this, can actually become part of the law making process, an act of discourse in many different ways. 

In this day, standard setting organizations are ‑‑ international standard setting organizations are often more influential than any one government.  So we see a more complex set of act interactions between actors which I think is much more interesting than asking what government's level should be. 

>> MODERATOR:  Bob.

>> BOB PEPPER:  The question was about standards; and, actually, if you look at some of the most effective standards, governments have had very little, if anything, some cases nothing to do with them, like IETF standards, the technical standards, or even the standards coming out of 3GPP which created 3G, 4G mobile service.

So I do think that actually standards development, standards enforcement is largely on the technical side multistakeholder.

If we're talking about enforcement against bad behavior as I said earlier, at the end of the day, there needs to be a government role in enforcing actions against bad behavior, whether it's through a private mechanism or a government mechanism. 

>> MODERATOR:  Eduardo.

>> EDUARDO BERTONI:  If we believe that the government should have some role, the role should be based on some previous regulation.  And the context example that you gave here when FCC stopped, it stopped because there was something before, some regulation that allows the FCC to act.  If not, it's impossible for FCC to act.

So if you're saying, well, FCC at that point did something, even without having proper net neutrality regulation, well, I agree; but there was something that could ‑‑ well, there was something that was the basis to act. 

So at the end, we are believing in regulation. 

>> MODERATOR:  Harold. 

>> HAROLD FELD:  What we mean by standards is another one of these exciting questions.  If we are talking about technical standards like IPB6 verses conducts of standard and conduct of behavior, I think that government has generally been less successful at choosing specific standards.  Although I can point to cases where the choice of standard was somewhat arbitrary, but there needed to be one uniform standard ‑‑ that digital television conversion is such a case, for example. 

But the broader question here about the role here of the enforcement I think is exactly right. 

And I do not share Berin's distrust of telecom regulators.  I think expertise in this area is necessary.  Our difference of opinion derives in part, as do our theory of what happens in government and how government operates; and I will reserve that argument for another time. 

>> BERIN SZOKA:  Before we take the vote, I want to say two quick things. 

Number one.  At 2:00, we're having a discussion on the economics of deployment.  Please come to that.  It's in Room 8.

An two.  If anybody wants to talk about starting a Dynamic Coalition on deployment, come talk to me. 

>> MODERATOR:  Okay.  Vote time.  So when we first started, we ask you:  Net neutrality:  Yes, no, or maybe? 

So we want to hear back from you after this superfun panel and rock star panelists whether you change your mind.

So who says yes? 

Let's see the yeses. 

Okay.  How many? 

People are changing their minds.  Don't change your minds.  If you are yes, you are yes. 

18?  Don't change your mind.  Are you yes? 

Okay. 

>> AUDIENCE:  Yes.  Yes. 

>> MODERATOR:  20.  Okay.  We have 20 yeses. 

Who says no? 

Okay. 

No. 

Nine. 

Okay.  How about the maybes? 

>> MODERATOR:  Eleven? 

Eleven.  Okay.  When we first started, we had like ten yeses.  Now we made it to twenty. 

Congratulations, yes camp.  You have ten more supporters.

(Applause) 

>> BERIN SZOKA:  But ask who changed their mind or not voted originally. 

>> MODERATOR:  Okay. 

So when we first started, we had like six nos and now nine nos. 

So congratulations, too. 

But I think ‑‑ everyone left because we are ten ‑‑ so when we first started, we had like 14 maybes.  Now we only have 11.  So, sorry, but your panel has left early.  He wasn't a fighter. 

(Laughter)

>> MODERATOR:  So but before we leave, I want to ask you whether you enjoyed this panel.  Great. 

And we also enjoyed having you here.

And one last question.  Who changed their minds? 

Let's see the ‑‑ one, two, three ‑‑ only four brave people accept the fact that they changed their minds. 

Okay.  Thank you very much.  Enjoy your lunch. 

(Applause)

(Session concludes at 12:42.)