[NOTE: This is live blogging. It is not close to a transcript, nor is it comprehensive. Finally, I'm hindered by having only a sketchy sense of what they're talking about.]
Charlie Firestone of the Aspen Institute moderates a debate. Resolved that the Communication Act’s stovepipe, vertical regime ought to be replaced by a horizontal regime.
Rick Whitt (MCI and author of “Taking a Horizontal Leap Forward“)
Tim Wu (U of Va law prof)
Randolph May (Progress and Freedom Foundation)
James Gattuso (Heritage Foundation)
Rick Whitt: The basics of the Net are at odds with the Communications Act: Layers, agnosticism of IP, transparency of the layers (= end to end). The CA views it vertically: Title II covers voice, Title II covers audio/video, etc. Vertical regulation stifles innovation. In fact, a packet is a packet and thus upsets the silo-ization: an audio packet is the same as a video packet. The horizontal approach regulates by layer: content layer, application layer, transmission layer, physical layer. [I'm not sure I've gotten this right] and an intermediate one is consistent with the architecture of the Net. It also gives you more granularity.
May: The horizontal approach is based on techno-functional capabilities. But tech changes rapidly, so you don’t want to lock in public policy framework based on technology? We’re not going to have agreement on the layers. Fundamentally, this is not worth overturning the stovepipes to move3 to this place. We need to be in a better place: A regime that would look at services offered by providers in a market and to see whether those providers have market power, and if so what type of reglation you would apply based on market power.
Wu: The layers model need not be complicated. The one we’re proposing is the same as the model in the heads of the best FCC regulators. It’s based on the distinction between transport infrastructure and applications. Contol over the physical infrastructure restricts market entry. People are upset about this because the layers approach would remove the incumbents’ ability to block market entry. We need to control the physical layer because that’s where the bottleneck,
Gattuso: We agree that there’s a problem. But the Layers approach is muddled; people disagree about it. The key factors that should drive regulation: Competition and choice. Layers can inform you about what the market might be, but it’s not definitive. Competition should be the key consideration. Public policy should look at the actual problems we’re facing, not as a secondary consideration.
Whitt: May, the CA is not techno-functional. Yes, tech will change over time, but the layers have survived for four decades. The point isn’t to replicate the OSI stack but to give regulators a way to think about this. What we’re proposing to Congress is a two-layer approach. The main dispute goes to the broadband layer: MCI belives there’s concentration of broadband suppliers. Even Michael Powell thinks there’s an issue.
May: When you have a model that leads things unclear, you’re inviting litigation. The FCC is trying to look at services to see if they’re in the same marketplace. That doesn’t involve a technical-functional distinction.
Wu: Any telecommunications legislation will have classification. The point of layers is to minimize them and have them make sense. The vertical model muddles the question. Would a market or anti-trust approach be simpler? No, you have the same problem of market definition with anti-trust. It’s even more complex. Leaving it to anti-trust is just a way of saying that we should just leave the incumbents alone. Anti-trust courts rarely do anything.
Gattuso: Regulation ought only to look to whether competition is working. We’re actually talking about whether cable companies have market power over consumers. That’s what the discussion ought to be about. I worry that with Layers model in 15 years we’ll be arguing over the layers instead of over the real question which is whether consumers have choices. And when Rick says “it’s hard to imagine” that tech will change, that’s exactly the problem we’ve had with communications acts.
May: Rick may be coming over to our side. In his new handout he says that all entities should be free to compete within and across layers without regulation. I don’t think Prof. Wu agrees. If this is nothing but a market power test, that’s what James and I are saying.
Firestone: There’s agreement that: A. The existing regime is too restrictive. B. Extreme market power is bad and we want some kind of anti-trust. C. We want to get to a place where there’s more competition and consumer choice. Whitt says we should get there by changing the scheme so there are basically two layers. Gattuso wants a strict competitive approach, getting more competitors into the market.
Wu: If we all agree on an anti-trust framework and principles, then you realize that the Layered model deals with a repeated anti-trust issue, i.e., the abuse of the physical layer to restrict competition. It’s a way of dealing with a repetitive anti-trust problem. People who believe in anti-trust principles ought to be on our side.
Frankston: Layers are a great talking point but there’s no reality to them. We’re starting with the assumption that regulation makes sense. Maybe we should recognize that provisioning bits isn’t a good business.
Whitt: The regulators don’t think about it as bit pipes. The Layers approach tries to shatter that way of thinking. It’s about political feasibility.
Tim Denton: If we put the four of you in a room, you could come up with the right legislation.
May: Yes, but our side wouldn’t put in language about layers.
Isenberg: Can Randolph and James put forward some simple principles to make sure we get the best network, given today’s reality of big honking companies that have captured the regulatory apparatus.
Gattuso: I’d make sure that one set of big honking companies can compete to provide alternatives; you want them to go after each other. I’d make sure people can get a foothold in the market; that may be in conflict with uniform connectivity and universal standards. I tend to favor approaches where the regulator is more general, e.g., Federal Trade Commission, not the FCC.
Richard Levine: In the European model, you define anti-trust markets and whether a single firm dominates; if so, then you do something about it. E.g., UK has a market called “broadband access” and British Telcom has a dominate position.
Joel Plotkin: The Baby Bells have privatized a public asset and are blocking competition.
Jerry Gleason: You’re talking about competition but you’re funded by the incumbents. [They disagree.]
Fred Seignor: How long will the layers model survive with Verizon buying MCI.
Wu: At a conceptual level, anti-trust and European model are attractive. The question whether practically they are excuses for doing nothing. The MCI 2-layer model is real legislation to combat abuse of the physical infrastructure.
May: If you believe that generally that competition is better than regulation, and that you have to provide incentives for people invest and innovate. You do that, you don’t take over their property. Just by calling them “incumbent,” that’s not a policy. [Technorati tags: f2c mci]
Categories: Uncategorized Tagged with: conference coverage
Date: March 31st, 2005 dw