In the House Judiciary Committee several members have proposed making network neutrality an element of anti-trust law. This is an interesting approach, but perhaps a bit heavy handed. Granted, I have yet to read the text of the proposal.
Tags: Net Neutrality
This is a very interesting post from the Progress and Freedom Foundation speculating that the argument for network neutrality presents a postmodern conundrum. Im not sure I go the same places the author Ray Gifford goes with his post.
The PFF have become synonymous with anti-network neutrality side of this debate, and the think tank tends to have a liberterian bent to their arguments. I have a lot of respect for their philosophical foundations, and in fact PFF Senior Adjunct Fellow Solveig Singleton was effectively my boss when I interned at the Cato Institute during my senior year at GW to research telecommunications policy. Solveig is an excellent thinker and I learned a lot from her in my time at Cato.
In general I think the folks at PFF who make a free market argument for network neutrality are ignoring the anti-liberterian history of blatant rent seeking on the parts of cable and telecommunications companies. They’re not evil– they’re just not the supporters of laissez faire that they are trying to present themselves as at this point in time. I really wish free market thinkers could see this, but I digress.
Mr. Gifford’s post makes the point that requiring a neutral Internet opens up some possible paths of innovation but closes others. This is certainly true, and the neutral network approach certainly favors dumb, standards-based, interoperable, best effort central networks. Such a mandate would close off a potential path of innovation in which a smart, proprietary central network with tiered classes of service would be able to offer. Mandating network neutrality, he argues, is a display of a normative preference for one innovative path over the other. The post goes on to ask why such an open innovative path should be preferred over a proprietary, tiered path. This is a very good question.
The simple reason the neutral network is better is that the open path opens up the field of innovation to many more parties than the proprietary path does. Gifford posits that a tiered approach opens up more paths to innovation, and I simply do not agree. Consider how much innovation in residential telecommunications services there was in the many decades before the breakup of AT&T, in the period between 1984 and 1996 and then again following the telecommunications act of 1996. If you look at each one of these historical period you see that innovations in service are opened up at each step when control of the end-to-end infrastructure is segmented. As this happens, standards become more important, and new devices enabling a growing number of services are placed in the increasingly fragmented standards-based network. Now, if you compare the amount of innovation that has taken place over the Internet, the ultimate open standard network, versus the switched telephone network there is absolutely no comparison. Clearly the dumb network model makes greater innovation possible, increases competition and encourages use of the network and services by end users.
Mr. Gifford closes his post by stating:
I just can’t figure out how net neutrality proponents can be so assured of the normative superiority of their outcomes.
Proponents can be assured because history has proven this to be the case. A dumb network will enable new content and application providers. Clearly innovations in compression, caching and distribution have been able to overcome the limitations associated with the best-effort nature of Internet traffic. This model also ensures there will continue to be content and services which appeal to end users, which ensures they will keep purchasing broadband service. Net neutrality is clearly a case of a rising tide lifting all boats.
Tags: Net Neutrality
There is a GREAT piece written by Daniel Berninger posted on GigaOM that talks about how telecommunications carriers may already be beholden to state and local requirements that would limit their ability to charge premiums for Internet traffic. These limitations would be based on the requirement that their services be offered on a common carrier status in exchange for no-cost or low-cost access to public rights of way to build out their telecommunications facilities. Some great opinions on both sides of the fence were shared in the comments page as well.
From the piece:
The obligations established on a state by state basis sometimes include build-out requirements or other compensation, but they all specify that access to state right-of-way at largely no cost or limit requires common carrier status (aka net neutrality.) The loss of common carrier status invalidates the contracts. The Bell companies have no access to state right-of-way for deployment of private, closed, non-neutral, non-common carrier network deployments.
This is one of the arguments that Ive been making for some time now, and is a solid libertarian argument for network neutrality. These companies have built their capability through exclusive access to government largess. This compromises the telcos claims that their property rights over their infrastructure allows them to effectively do and charge whatever they want with those pipes. My argument is not one for heavy handed government oversight (federal or otherwise) of the use of that infrastructure, but clear statements from regulators with respect to their general expectations on the treatment of traffic from multiple content and service providers over this infrastructure. A little humility on the part of the large telecommunications carriers wouldnt hurt, either.
Another argument Ive made several times is that telcos and ISPs get all sorts of legal passes on questions about the transmission of pornographic material, illegal file sharing and libelous e-mail and web pages based on the claim that they are a common carrier. By making exclusive arrangements with content providers or discriminating in the transmission of packets the telecommunications companies compromise their claim to be a common carrier. Since they are now discriminating one kind of traffic from another, it doesnt seem like a significant stretch in logic to conclude that they could or perhaps should be held liable for child pornography or illegal copies of music files which move across their networks.
On a slightly related note, yesterday at a telecommunications networking event I met someone involved in working on broadband over powerline equipment. Its good to know that new facilities will soon challenge the cable and teclo duopoly with new competition for broadband service.
Tags: Net Neutrality
Om Malik has written an excellent blog entry on network neutrality. It’s definitely worth a read.
The pursuit of tolls based on content and application type requires something that does not exist in the Internet today. It requires a linkage between content type and transport. Equipment providers like Cisco increasingly deliver products offering packet by packet inspection in the name of network management, but implementing the access fees means giving billing systems the ability to monitor and track the types of applications and content customers use. Setting aside the chilling privacy concerns, the telephone networks linkage of usage to transport represents the primary obstacle to service creation I observed during five years at Bell Labs in the 1990s. Forcing innovators to change the network in order to implement an application means an end to innovation. The end of innovation means the end of growth in demand for Internet access.
The focus of Om’s entry in on the threat of innovation to incumbent broadband providers’ traditional revenue streams. I think his comment on his time at Bell Labs is telling, because it goes back to the net-head versus Bell-head debate of smart versus dumb networks. In the economic model that the Bell companies have been proposing is the resurgence of the smart network, where the Internet has traditionally been the domain of a dumb network with very smart applications and end devices.
While a “smart” network sounds like a very nice thing, what it really represents is a model for application delivery which relies on major revisions to network infrastructure. This comes through expensive, proprietary development and limited market access. In this model the incumbent network provider controls any new products or services, and those will be rolled out at a very slow pace and at very high incremental cost to subscribers. Look to the first implementations of call waiting, voice mail and caller ID as examples for the telco model of rolling out value-add functionality. It’s not pretty.
This proprietary model flies in the face of a number of trends in IT that have been influenced by the open standards-based environment that is the Internet. Think open source, think APIs, think enterprise applications based on off-the-shelf software tools from ERP vendors. All of this replaced models where very large teams of developers worked very hard at great expense for very small incremental improvements.
Hopefully Congress will look favorably on the concept of network neutrality and clarify the situation.
Tags: Net Neutrality
Finally, after at least two months of promising, Im ready to give a review of the second draft of the telecom law currently being considered in Congress. This comes just in time for hearings on the subject on Tuesday of next week.
I did happen to read a story this week that Verizon and the other RBOCs are trying to convince Congress to move away from considering comprehensive telecom policy reform in favor of piecemeal reform. Among the other issues the RBOCs are eager to consider on a piecemeal basis is clarification that they should not be required to seek franchises from local franchise authorities.
My first reaction to this draft of the bill is that without specifically mentioning the OSI layers the language of this draft does a much better job of conveying the differentiation among the layers involved with delivering a telecommunications service. Especially noteworthy is the differentiation of the layers involved in delivering Broadband Internet Transmission Services (BITS) and a decent legalese definition of packet-switching. The bill also includes a definition of ILEC, which is good for a change.
Part of the improvement in this draft is the new clarity with respect to the definition of Broadband Video Services (BVS). One clarification is that BVS is a two-way service that integrates video programming and data services. Despite the clarity that this provides, the language of the bill does not specify the means by which a traditional cable companys services can be transformed into a BVS service. This lack of language is a bit off considering the bill does include very clear language that does indicate a cable company can make that transition. Considering BVS providers have more freedoms from franchise requirements than traditional cable companies this is a big deal that is going to be litigated if future versions of the bill dont clarify the language.
The bill taketh away and giveth as far as BVS are concerned. Service providers will not be required to get authorization from local franchises. The local franchise authority retains the authority to charge franchise feesallegedly on a non-discriminatory basis. Most troubling the local franchise authority retains right to define the fees, schedules and general say-so for access to rights of way and easements for the installation of facilities. I still think this is going to be the long pole in the tent for true competition of any kind so long as the RBOCs and incumbent cable companies have pretty much exclusive access to easements.
With respect to VoIP, the limitations of requirements for Universal Service Funds and E-911 seem to be limited to only those services that offer dialing to telephones whose numbering defined by the North American Numbering Plan. This language seems to exclude pure PC-to-PC calling, which is a way to avoid an enforcement nightmare. Also, receive only VoIP services are excludednow, if only we could know exactly what that means.
The FCC is empowered by the bill to consider whether the migration away from traditional land lines will necessitate the extension of USF requirements to VoIP services. This seems to me to be pre-concluded that the answer to this question is going to be a definitive hell, yes! from the FCC. Im still somewhat disappointed that the bill doesnt provide language that allows the FCC to consider basing the USF contributions on telecommunications facilities as opposed to services oreven betterto consider the question of whether or not the USF still has value. Considering how many municipalities are interested in deploying their own fiber or Wi-Fi the USF appears to be of questionable value in this day and age.
Interestingly, VoIP providers are going to be required to support number portability. No doubt this will be applied with all the same grace and finesse (note the sarcasm) of the requirement for 911 and E-911 capability for VoIP providers. Expect this to be a hammer to fall if any major VoIP providers find themselves challenged with this technical development. Hopefully lessons will be learned from the early problems with number portability among wireless carriers.
Overall the language of the bill is getting better, but there are still a few things that one wishes Congress would reconsider. Also, if were going to have a comprehensive telecom bill it would be nice to have some very specific language with respect to network neutrality. Considering the recent comments of executives at three of the four RBOCs it seems Congress would do us all a service by clarifying the issueideally in favor of a definition to ensure content from competing companies would not deliberately be discarded.
I came across this piece on the Networking Pipeline blog (thanks to this entry on Techdirt).
Google’s {PR Spokesperson] Barry Schnitt told Paul in an email: “Google is not discussing sharing of the costs of broadband networks with any carrier. We believe consumers are already paying to support broadband access to the Internet through subscription fees and, as a result, consumers should have the freedom to use this connection without limitations.”
It’s great to finally here some (muted) response from the other side of the argument over network neutrality. Of course sending a big “F you!” to the Bell companies doesn’t mean anything if you can’t play the regulatory game the same way that they do.
On a somewhat related topic, I’m very slowly making my way through the second draft of the telecom bill, which was released way back in November. The legal language has holes wide enough to accommodate a speeding semi tractor-trailer and ensure that Internet content and service companies will be in litigation with Bell and Cable companies for at least a decade after the bill actually passes (if it passes in anything like its current form).
An executive from an RBOC sounds off on Internet applications not paying their fair share:
“We have to make sure that they [application providers] don’t sit on our network and chew up bandwidth,” Seidenberg said. “We need to pay for the pipe.”
These are the words of Verizon CEO Ivan Seidenberg at CES according to this article from TechWeb.
The article continues…
“We would not do that [block specific applications], and it shouldn’t be done,” said [Executive Vice President for Public Affairs] Tauke, who said that Verizon is actually in favor of codifying so-called “network neutrality” rules, as long as there are provisions for parallel private networks, like Verizon’s nascent fiber-based television services. At CES, Verizon announced plans for an online gaming network that would also likely be segmented away from the “open” Internet, so that Verizon could better control the performance of the network.
The paralell network mentioned by Tauke could reflect a similar concept to how cable companies today seperate out the delivery of television content and broadband Internet. The benefit of this is that it would ensure the throughput of both services. Cable television is not delivered over the same layer 2 network that delivers broadband cable Internet, and RBOCs offering television services should certainly be free to deliver their television service in a similar manner.
I’m still a little concerned about the comments that application providers “chewing up bandwidth.” My major problem with comments like this is that they continue to ignore the fact that each customer pays for their broadband service. Likewise application providers pay for their connections to the Internet. In each instance, each customer can only be expected to get what they pay for, an amount which is hopefully communicated upfront in terms of a bandiwdth capacity or amount of traffic I can download. If I purchase a certain broadband capacity from my cable company or RBOC I should be able to fill up my alloted capacity with whatever it is that I want. Presumably this also gives me a claim to some (very small) percentage of the total capacity of the broadband provider’s backbone network. If it all comes from Google, Yahoo, or AOL then who is to complain about that? If my net usage causes a problem for other subscribers then it is up to the broadband provider to throttle back my capacity for all applications and services, not to do so discriminately based on who has paid a premium. Of course all of this throttling should be done in a transparent manner, as well.
I must have been just busy enough with work and whatnot this week to miss this. This evening I caught an entry on Om Maliks blog a few days ago commenting on the double standard that the FCC is evidently applying between cable and incumbent telecom providers. Om refers to an article in the Wall Street Journal summarizing a request for information sent by the FCC to Comcast and Time Warner as part of the review of their pending acquisition of Adelphia.
From the Wall Street Journal (subscription required):
The FCC’s information request also focuses on several delicate areas, including agreements for regional sports programming and “net neutrality” rights — essentially preventing companies from discriminating against Internet traffic. The latter request is notable because FCC Chairman Kevin Martin wasn’t particularly concerned about net neutrality in two recent Bell mergers.
The segment of the original FCC request to which I believe the Journal is referring is as follows:
Provide all Documents relating to deliberations and decisions to stop, limit, hinder, slow, or otherwise impede the transmission of information over the Companys Residential High-speed Internet infrastructure based on the software application, source, destination, or other characteristic of the traffic. Documents regarding unsolicited commercial e-mail and malicious software need not be produced.
I suppose on some level we should be relieved that the FCC is paying attention to the issue of network neutrality. There are still two outstanding issues, though:
From Reuters:
“I’m hesitant to adopt rules that would prevent anti-competitive behavior where there hasn’t been significant evidence of a problem,” [Fcc Chairman Kevin] Martin said at a conference luncheon by Comptel, a group representing competitive telephone carriers.
“That doesn’t mean people don’t have a lot of concern about potential problems, but there’s a significant difference between potential problems and problems that occur,” he said
Well, crap. While I can’t disagree with the rhetoric that potential problems and real problems are not the same thing, the recent comments from AT&T and BellSouth executives make it pretty to clear to me that there was a need for some sort of clarification on this subject from the FCC. This is not what I would have hopped the Chairman would articulate.
Hopefully Congress will clarify this situation better.
According to this post on boston.com, AT&T and BellSouth are lobbying on Capitol Hill to establish a premium tier of the Internet. Based on the text of the article the nexus of the argument is that the Bell companies need to offer a premium tier to deliver broadcast video services. This is further indication that the Bell companies barely understand their own business and have even less of an understanding of how video services are delivered. Either that or this is an effort to confuse members of Congress.
The telecom companies said that since they are spending billions of dollars to build new fiber-optic networks that can carry more data, they are entitled to give their own offerings the bulk of Internet bandwidth, and to charge others for higher-speed access.
”When costs are being driven into an equation, they have to be recovered somewhere,” said Bill Smith, chief technology officer of BellSouth. ”Why do fundamental business economics not apply to the Internet?
The Bell companies are arguing that they need to offer a premium tier in order to deliver their own video services. I have to contrast the claims of the Bell companies with how the cable companies delver video. In the case of the cable companies video services are delivered using a very different infrastructure than Internet services, and the only thing video and Internet services really share is the pipe into the house, and only then at the physical layer (OSI Layer 1). The two services occupy different electronic streams coming down that pipe and never the twain to meet. This is a very real simplification, and I readily admit that my knowledge of how cable television is delivered is about an inch deep.
The Bell companies are arguing that they need a premium tier for their own video to offer television services. I dont know enough about how the Bell companies plan to deliver their video services over fiber, but I would be very surprised if they were going to deliver it over the same wavelengths (this is fiber were talking about) as their Internet service. Even if the Bells are using the Internetworking Protocol to deliver video from the local head-end to the home, it would make no sense for that delivery to share the same network space as broadband Internet service. If these two streams did share any part of the network beyond the physical connection itself then your Internet browsing would grind to a halt every time soneone in your neighborhood turned on a television.
Does anyone believe that the Internet is really ready to deliver full-screen, real-time video? Multiple video programming to each subscribing household? To multiple households in the same neighborhood? I would be really surprised if the Bells planned to deliver video and Internet over anything more than the same physical connection into the home (again, OSI layer 1 only). If the cable companies can manage to deliver video and broadband Internet without handicapping third-party Internet content then I dont think its asking too much to ask the Bell companies to manage the same.
What I suspect is going on is that the Bell companies are taking advantage of legislators lack of understanding of how telecommunications services are really delivered (and journalists’ lack of understanding from the look of articles I’m seeing on this subject) to make a land grab for control over all of the OSI stack based on their duopoly for layers 1, 2 and maybe layer 3. (Yeah, I still need a post explaining what the hell all this OSI layer stuff means).
augustjackson.net RSS feed Competitive Intelligence Podcast
This blog is offered under a Creative Commons attribution share-alike license.
