Controlling the 'Net: How Vested Interests Are Enclosing the CyberCommons and Underming Internet Freedom

An Interview with Lawrence Lessig

Lawrence Lessig is a professor at Stanford Law School. He is the author of The Future of Ideas: The Fate of the Commons in a Connected World and Code and Other Laws of Cyberspace. He clerked for Judge Richard Posner on the Seventh Circuit of Appeals and Justice Antonin Scalia on the U.S. Supreme Court. He is a monthly columnist for The Industry Standard, and a board member of the Red Hat Center for Open Source.

Multinational Monitor: What do you mean in describing the Internet as free?

Lawrence Lessig: Think of the Internet as broken into layers. There is the physical layer, which are the computers and wires that make the Internet run; the logic layer, which are the protocols that make it function; and the content layer, which includes the programs that run on it — MP3, for example. When I say the Internet is free, I’m referring to the middle layer, which is unowned and embraces an architecture that ensures that people can innovate for this network without the permission of anyone else. But obviously, the physical infrastructure — the bottom layer — is owned by people, and much of the content is owned and controlled, so those parts aren’t free. But the core is free.

MM: You suggest that things that are owned may in some cases also be understood to be free.

Lessig: Right. One central idea in the book is the concept of the commons, which I define as a resource that is either completely unowned or whose access is granted to people equally. For example, if you charge people a dollar for admission to a park, it functions as a commons because you’ve charged a price equally to people regardless of what they want to do with the park.

Common carriage regulation in that sense takes what is otherwise owned property, like a telephone system, and makes it function like a commons, because there’s no discrimination in access to the resource.

At the physical layer, the original Internet was layered on top of a telephone system that was regulated to be free in that sense: under common carriage regulation the phone system was regulated to make resources — the telephone lines — available to others on an equal and neutral basis.

MM: When you talk about the code layer of the Internet, one of the central ideas is end-to-end. What does that mean and why is it important?

Lessig: The core architecture of the network is “end-to-end.” That means that the network itself is designed to be as simple as possible. All the intelligence is placed at the end, or edge of the network.

Peer-to-peer is one implementation of this network. What it means is that computers at the edge of the network are communicating directly with each other, not through a server in the center of the network or some central computer which is allocating resources to the edge.

Napster was a kind of peer-to-peer system. You were essentially transferring files from one person on the edge of the network to another person on the edge of the network.

Instant messaging technologies are peer-to-peer, because one computer at the edge of the network is talking to another computer at the edge of the network.

These systems importantly are not controlled by anything built into the network.

MM: The central theme of your recent book is that some of these core aspects of how the Internet has been free or decentralized are under threat. To what extent is the physical layer, including the existing telephone wires and DSL and cable, open, and what are the trends that seem to be dominating in that physical layer?

Lessig: As the Internet moves from narrow band to broad band — from telephone lines to cable lines — cable companies have the power and the right legally to exercise much more control over what happens on the network, because they are building and deploying technologies that will enable discrimination in the content and applications that run on the network. That’s the kind of discrimination that the original architecture of the network didn’t allow. That’s solely a function of moving from telephones to cable, and an accidental feature of regulation — which is that cable has no mandate to be neutral and open, whereas the telephone companies did.

MM: What kind of discrimination might the cable companies be able to exercise?

Lessig: There is a technology called policy-based routing, for example. Policy-based routing is implemented through a router that allows the cable owner to choose which content flows quickly, which content flows slowly, what applications are permitted and what applications are not. It’s like a television set where channel 6 is clear and channel 12 has static on it, and it just so happens that the television manufacturer also owns channel 6.

What this means is the platform of the Internet, which is what you expect to get when you connect broadband, is no longer neutral among the different uses of broadband technology. Instead, it’s tilted in favor of some content and some applications, and against others. It’s that compromise of neutrality that is going to create a great burden for innovation and creativity.

MM: Is there evidence that the cable companies would use these kind of technologies?

Lessig: First of all, they’re deploying them now. Secondly, they’ve exercised this kind of control with respect to video across the network. Yair Landau, the head of Sony’s broadband technology, described at a Stanford event how cable companies had told him that if Sony delivered pay-per-view movies across the Internet, then cable companies would shut down the Internet before they would ever allow that on their system.

So the answer is yes, they certainly have expressed the desire and are already deploying the technology to enable that kind of discrimination.

MM: To what extent is DSL, a high-speed broad-band service delivered over the phone lines, an alternative?

Lessig: It’s competitively an alternative in the sense that phone companies typically don’t have content bundled with them. So they have less of an interest in playing this kind of game.

But they’re an inferior technology in the sense that it’s going to be much harder for DSL to provide the kind of throughput or power that cable can. Cable Labs, which develops the cable modems, now has a cable modem standard that allows 40 megabits per second in both directions. This is version two of their cable modem. 40 megabits per second is extraordinarily fast compared with DSL, which is lucky if it provides 1.5 megabits per second.

So in the long run, cable is going to be a much more powerful platform to deliver broadband content. The fact that telephones right now are a better competitor doesn’t guarantee that cable doesn’t get the lion’s share of the market in the long run.

MM: Are there phone companies or other vested interests against the cable companies exercising this kind of control?

Lessig: Because of the regulatory history, phone companies are required to be open access, which means they are required to allow competitors to use their phone lines. So if you wanted to set up a Internet service provider to compete with the phone company, they’d have to let you do that. Cable companies don’t have to let you do that, because cable companies live under no similar regulatory burden.

As a result, phone companies have pushed for a long time to try to put cable on the same regulatory plane that telephones are. So they’ve been strong opponents.

I also think that, in the long run, Microsoft is going to be a strong opponent to these games being played on the network. Microsoft’s strategy for .NET envisions the network remaining the way it was. Microsoft wants to deliver services to the edge of the network without having to worry about what’s going on in the middle. So they could become a powerful ally in defending the original architectural values of the network, but it’s a little early to know.

MM: In that hypothetical, is the Microsoft concern that it wants to be able to distribute efficiently, or that the cable companies will actually rise as a competitor to Microsoft?

Lessig: The cable companies could become a kind of competitor to Microsoft, especially if they get allied with content and AOL.

AOL is becoming an increasingly vertically integrated entity — Time Warner is the largest content-owner, and AOL obviously has a huge percentage of the online access through their proprietary software, and they increasingly own cable lines. So they own content, code and the physical layer of the network, and that integration creates barriers that make it extremely hard for competitors to battle with them. So that will be one dimension of the battle that Microsoft has with them in the future.

MM: How did the AOL Time merger affect AOL’s position on this set of issues?

Lessig: AOL Time Warner made AOL the single most powerful competitor to Microsoft. It’s also possible that this merger made AOL dangerous for the future of the Internet. It’s not clear yet, but this vertical integration creates all the wrong incentives for keeping the platform of the network open.

This question of whether open access should be pursued as government policy has been answered in different ways by both AT&T and AOL, depending on whether they happened to own cable lines. When AT&T was in Canada before they owned cable lines, they told the Canadian government that it was extremely important that open access be required for cable lines, because otherwise the cable company would exercise too much power over the Internet. Once they owned cable lines, then they said the market should decide this, not government. Same thing with AOL. When AOL didn’t own any cable lines, they came to a number of governments, including in San Francisco, and said they needed to impose open-access requirements. Once they owned the cable lines, following the merger with Time Warner, they were much more willing to let the market take care of itself.

MM: You talk about copyright in the pre-Internet era as embodying a kind of balance between competing interests. What competing interests did that balance serve and how did it play out as technologies evolved pre-Internet?

Lessig: Copyright has always been understood to be a balance between incentives to create and access of the public to the things that are created. The notion was, once you’ve created the incentive to produce something new, as quickly as possible you want to pass it into the public domain so that other people can build on it in a free way. The framers’ [of the U.S. Constitution] conception of copyright initiated that move into the public domain very well, because their conception of copyright had a very short term — basically 14 years, renewable once. And they applied it to a relatively narrow range of rights, basically protecting only maps, charts and books against re-publication. That was their striking of a balance to ensure there was an incentive to produce while guaranteeing that creative works would pass into the public domain after a short interval.

Now that balance has changed. Increasing copyright protections have occurred, and largely for good reasons. A lot more is protected. But at some point that protection becomes too strong. The balance is no longer a balance. That’s where I think we are now, because it’s not just that the term is extremely long — the life of the author plus 70 years — but the scope of the right is extremely broad. It covers copying, derivative works, using the works in different media. All of this control in the hands of the original copyright owner can inhibit follow-on innovation. This is particularly true of digital technologies, which enable that kind of innovation.

MM: As new technologies evolved pre-Internet, how were conflicts between providing protections and facilitating new innovation resolved?

Lessig: There’s been a pattern where when a new technology comes along that changes the mix between protection and access, for Congress to deal with this through something like a compulsory licensing regime. Compulsory licensing is a regime that gives the new technology access to the copyrighted material, but makes sure that the new regime pays for that access.

The first example of that is the player piano. Before the player piano, copyright owners of music made their money by selling sheet music. The player piano made it possible for makers of player pianos to Napsterize the sheet music market by buying one copy of the sheet music, converting it into a piano roll, and then selling the piano rolls — thereby making it no longer necessary for people to buy sheet music. The U.S. Supreme Court held that that was not a copyright violation.

So Congress had to regulate through a change in the statute. When they changed the statute, they gave the sheet music manufacturer the right to make copies after an initial mechanical production has been made and pay a fixed rate for those copies, like two cents per copy. That was a way to make sure copyright owners got paid, but the sheet music industry couldn’t leverage their power over sheet music into control over the next technology, which was player pianos.

The same thing happened with cable television. Cable TV is the most dramatic case of Napster before Napster. They set up technologies — television receivers — on top of hills, stole the content of broadcasters and sold it to their customers, delivering broadcast programming by cable. Twice the Supreme Court was asked to strike this down, and twice the Supreme Court said it was not a violation of copyright law. For 20 years, the cable television industry got to Napsterize broadcasting content without regulation of copyright law. When Congress got around to regulating this, they struck the same balance that they struck in the context of player piano rolls. They said that cable companies have the right to get access to broadcast television content, but they have to pay for the content. So Congress created a compulsory licensing right with respect to broadcasting content.

In both cases, the objective is to make sure that one industry doesn’t leverage control into the next generation. That’s what a compulsory license right does. The same opportunity could exist in the context of the Internet, if Congress would ever get around to protecting the Internet from overly strong copyright regulation.

MM: Except that the Napster case came out the other way, and the court ruled that Napster was violating copyright.

Lessig: Partly because copyright law itself had been so broadly expanded, the court felt it could not conclude that this use of content could occur without the permission of the copyright owner. So the growth in copyright law meant that there was no doing-nothing option here. The copyright law would shut that down relatively quickly. And that’s the problem, because essentially what this means is that any new technology doesn’t just have to defend itself in the marketplace, but in the court room against copyright owners before it’s allowed to exist in the marketplace. That seems to me a good way to stifle innovation.

MM: The background law has evolved so that the presumption is to shut down the new technology. That changes the dynamic in the regulatory system and Congress altogether. Now Napster is the one who has to muster the political will to get a compulsory license, instead of the recording industry.

Lessig: Yes, and I think that tipping of the dynamic is devastating for this type of innovation. If you look at the way copyright law has been considered historically in Congress, leaving new innovation the opportunity to develop and then prove itself before you attempt to regulate it was a critical part of the balance. So player piano got to exist and take off long before any court tried to shut it down. No court did shut it down. It was Congress who had to regulate it. The VCR was the same thing. After it took off, the movie industry tried to shut it down via legal regulation, which failed, so then Congress got to consider whether there was something they wanted to do to deal with the VCR. Cable TV was the same thing. These are cases where we’ve been allowed to innovate first and regulate later. But the law now forces us to regulate first, and innovate later. “Later” in the sense a teenager would mean, i.e., never.

MM: What impacts are patents having on digital technologies?

Lessig: In the context of software and business methods, patents have exploded in the last 10 or 15 years. There has not been a lot of recognized patents in these two areas until recently. This changes dramatically the dynamic of innovation in both fields, especially the software context.

Software projects have to constantly monitor what they’re developing to make sure they don’t cross somebody else’s patents, so much of the money now gets spent on lawyers rather than engineers.

In my view, with any regulation, the person promoting the regulation should demonstrate first that the regulation is going to do some good, and then, once there’s some demonstration that the regulation will do some good, then you roll out the regulation. You don’t regulate first and ask questions later. But the problem that we’ve had in the context of patents is that the patent bar has pushed this form of regulation and the court of appeals for the federal circuit court has endorsed it, so we’ve seen this explosion of software and business method patents, without any good demonstration that this type of regulation will do any good.

In fact, there are lots of good reasons to believe it will do harm. We know that patents are hugely expensive to administer. It’s a dead-weight loss, economically, to society. With innovation like in software that comes in sequential, complementary fashion, there’s a likelihood innovation will therefore be hampered by patents.

What the policymakers should be doing is deciding whether there’s a good reason to believe the benefit from patents in this area outweighs the costs. But if you look at the economic literature, there’s no good evidence to suggest that the benefits outweigh the costs. That leads to the conclusion that we shouldn’t be regulating here.

MM: In the area of the physical layer of the Internet, what would be the alternatives to the control the cable companies are exerting?

Lessig: One thing is to expand the competition. I think wireless could be an important new competitor in this field. So we should support a much broader space for unlicensed spectrum and encourage innovators to develop that space by guaranteeing there will be unlicensed spectrum for internet transmission for the long term. That would create a real kick-start to the competitive process. The existing duopoly between telecom and cable companies (even there, only where the technology overlaps) creates weak competition. What you need is a new competitor, and I think wireless can do that.

To the extent that wireless can’t do that or it would result in insufficient competition, another solution would be to impose a common carriage regulation at the code layer. Not at the physical layer — this regulation would say, “You can build whatever wires or technology you want, you don’t have to open your wires up to competitors. But if you’re delivering Internet content, here are the rules under which the content must be delivered.” That would create a kind of neutral platform, imposed at the code layer.

Thirdly, you can impose open-access requirements at the physical layer, so that you induce competition to guarantee that strategic behavior by the cable or broadband companies could succeed.

In the content layer, I think there should be radical change in the regulation of copyright law — not abolition of copyright law, but a change in its scope and certainly duration to enable a lot more content to exist in the public domain. A certain amount of compulsory licensing rights would also free up lots of content for innovation.

MM: Linus Torvalds, who has been central to developing GNU/Linux, has recently published a much more optimistic book than yours. He thinks open-source code — which puts new innovations, like the GNU/Linux operating system in the public domain — has secured a permanent place for itself and that open-source innovation is thriving and here to stay.

Lessig: I hope he’s right, but I think he’s wrong. I think there are extremely strong interests allied against the neutral platform of the original Internet. I’m not sure that that’s Microsoft, but I am sure that there are businesses that don’t succeed in the world of open competition that the Internet produced and that they are acting to change that world.

It’s a separate question whether that means the end of open source. I don’t think it necessarily means the end of open source. But I also don’t think the survival of open source is a sufficient condition for the survival of the kind of innovation and creativity that I was describing.

I hope I am wrong. But if you look at where the politics lie, and what the interests are, the powers on the side of changing the Internet are much stronger than the powers on the side of preserving it, and that is why there is a reason to be skeptical.