Hide full article
In recent weeks a great deal of ink and pixels have been spent on the topic network neutrality. As set out by pundits and some of the mass media, a major battle is brewing between the big-money telecommunications companies who build the networks which comprise the Internet, and the big-money technology companies whose businesses depend on the Internet. The telecoms want to be able to charge Internet companies for preferred access to users on their networks; conversely, major Internet firms - and many consumer advocacy groups - argue such pay-for-good-access schemes will create an Internet of haves and have-nots, stifle innovation, and ensure only the highest paying online services are accessible.
In the United States, the issue has entered the political arena, with some members of Congress proposing to legislate network neutrality, and corporations and interest groups attempting to swing public opinion (and politicians' votes) their way. Propaganda sites for both sides have gone online.
<http://www.handsoff.org/>
<http://www.savetheinternet.com/>
So what's the brouhaha? Where did this debate come from?
Defining Net Neutrality -- It's odd to say, but as a practical matter the Internet has never been neutral. Access methods and capabilities have always impacted the Internet experience. Just ask anyone downloading Apple's 100-plus megabyte system updates via dialup: they'll tell you the Internet is biased against them! Similarly, you may have a broadband connection, but maybe your new favorite band's MP3 files are being served over an ISDN line and still download at a snail's pace, or perhaps your cable provider can't quite keep up with the streaming QuickTime movie. Or maybe your ISP allows you to send mail only via their mail servers, rather than through your employer's or one you run yourself. And firewalls - which are everywhere these days - are all about taking the "neutral" out of the Internet: they deliberately screen and block different types of Internet traffic. Corporations routinely block peer-to-peer file sharing and services which are known security risks; schools enable access only to approved "whitelisted" sites (or blacklist problematic sites: Del Mar College in Texas recently blocked access to MySpace). Some technologies "shape" Internet traffic by limiting how much of available bandwidth can be used by certain users, locations, or services, and, of course, some governments actively block and censor the Internet. These are all current examples of non-neutral behavior on the Internet today.
Further, Internet service has always been a pay-for-bandwidth privilege. There's nothing neutral about it: no one has a right to use the Internet for free. Although you might be able to skip between Wi-Fi hotspots these days and get decent Internet service without paying a dime, rest assured the folks running those hotspots are paying for that bandwidth. Connecting to the Internet costs money, whether it's via dial-up, cable modem, DSL, wireless, satellite, or another access method. Generally, the more money you're willing to pay, the more bandwidth you get. Today's Internet giants spend mammoth amounts of money to ensure they have bandwidth to meet their users' demands. You think Apple must have big Internet pipes to serve out those massive system updates? Think about Microsoft! Then think about the bandwidth that sites like Amazon.com, Google, MySpace, and YouTube must require. They aren't getting it for free, either.
Since in reality Internet access has never been "neutral," it's tough to define "net neutrality." However, at a basic level, you can think of it as representing the way the Internet operates in the U.S. (and much of the world) today. Leaving aside variables of access, policy, and bandwidth outlined above (and more besides), being connected to any portion of the Internet generally enables users to access darn near anything at best-effort rates. It doesn't matter whether you connect to the biggest online sites and services (like AOL, Yahoo, Google, MySpace, Amazon.com, or iTunes) or the most obscure blog, photo page, or teenage poem extolling Brad Pitt's abs. In the abstract, data from all those sites are treated equally and the Internet delivers information as efficiently as it can given the circumstances, regardless of who operates the physical networks the data may traverse. That's the essence of net neutrality.
Why Telecoms Are Worried -- The telecommunications companies that build major portions of the Internet - companies like Verizon, the reconstituted AT&T, Comcast, and BellSouth - are facing a dilemma. On one hand, demand for Internet and broadband services has never been higher, and it's growing all the time. On the other hand, the Internet-based services which broadband connections make possible erode the telecommunications companies' core businesses.
For example, the forthcoming capability to stream current television shows and first-run movies via the Internet doesn't bode well for services like Comcast's cable television business (or satellite-based TV services). Why subscribe to HBO if you can stream the movies you want over the Internet, whenever you want, for less money? Voice-over-Internet-Protocol (VoIP) services such as those offered by Vonage and Skype enable computer-to-computer phone calls via the Internet, along with video and conferencing capabilities, and can even call "out" to standard phones. VoIP and technologies entering the same arena (like instant messaging) strike at the heart of telecommunications companies' bread and butter: standard telephone service, often provided via exclusive franchise agreements. Telecoms are already losing (primarily young) customers who are dropping their landlines in favor of mobile phones. In the near future, they stand to lose even more customers to VoIP. (Particularly businesses: VoIP is attractive compared to business rate telephone and conferencing services, especially for international calls.)
And, just to make things more awkward, while demand for broadband services continues to grow, the amount of revenue a given amount of bandwidth generates for telecommunications companies has dropped as much as a hundred-fold in the last decade. Major bandwidth which would have cost millions of dollars a month in the mid-90s now costs tens of thousands of dollars per month. (This phenomena also scales down: my home business bandwidth bill is currently about 20 percent of what it was in 1996, and I have access to about 18 times the bandwidth.) True, the expenses telecoms incur to install that bandwidth have also declined, but not as quickly, and in some cases they're going up: acquiring right-of-way to put fiber in the ground is getting more difficult, and costs to upgrade hardware and construct mammoth networks are increasing, in part due to external factors like rising energy prices. Here's a quick way to summarize the rollout of U.S. broadband networks, even if it's over-generalized: the easy stuff has mostly been done, and from here on out expanding network capacity gets more expensive.
Defining Tiered Access -- The major telecommunications companies who create broadband networks look at their declining revenue growth and increasing costs, and see a not-so-rosy future for broadband networks (not to mention their revenue bases). And then they see the millions and billions of dollars being earned by companies utilizing existing broadband-enabled Internet services, and essentially say, "Hey, we built the networks those companies are riding on! Shouldn't we be getting a bigger slice of that pie?" It's even more galling when those Internet companies aren't paying much (or anything) for bandwidth on a particular network. Let's say Vonage buys all its bandwidth from Verizon (which isn't true, but play along). How do you think BellSouth might feel about Vonage offering VoIP service to BellSouth voice customers over the Internet, and maybe encouraging those customers to roll back or even discontinue their voice telephone service? All without Vonage paying to access BellSouth's broadband network? You can see why BellSouth might be upset, and at the same time, you can see why Vonage wouldn't see anything wrong with the situation; after all, they're already paying big bucks to Verizon.
Although that example is fictitious (most large-scale Internet services have to purchase bandwidth and specialized network access from several providers at many locations), the idea is simple: since all data transmission on the Internet is treated with more-or-less the same priority, Internet companies can use the telecoms' infrastructure to provide competing services and, in some cases, even steal their customers.
So, telecoms are looking for ways to generate and increase revenue from big Internet companies using their broadband networks. Additional revenue streams would mean more money for the telecoms, but also attract the investment capital necessary to build out advanced networking capacity and capabilities. After all, lenders and investors are mainly interested in backing projects which have hope of turning a profit. The question is: what can telecommunications companies offer the big Internet players which would sweeten the deal and entice them to pay - or, at least, pay more than they are now?
One idea is to create programs whereby Internet companies can pay telecoms to have their data treated preferentially. Say you run an Internet company that provides streaming video: what if you could arrange for your data to be delivered faster and more reliably over certain major telecommunications networks? Wouldn't that give you an advantage over your competition, keep your users happier, make you look better, and, thereby, improve your business? Wouldn't that be interesting to you? That idea is the essence of tiered access: for a fee, the telecoms could grant some Internet data a higher priority over other data.
Again, as a practical matter, Internet access already is tiered. In general, Internet access fees paid by companies and individual Internet users are proportional to the amount of bandwidth they receive. ISPs have been offering tiered access for years, encouraging users to upgrade from dialup to broadband services, and from "slow" technologies like ISDN and low-end DSL to "world-class" DSL, wireless, and cable modem solutions. Businesses face the same decisions on a larger scale, buying bandwidth, network access, and access to colocation facilities at varying rates depending on the services they need.
The Crux -- If implemented, tiered access schemes would change one of the Internet's fundamental (if abstract) principles: that all data is treated with the same priority, regardless of type, origin, or destination. Under tiered access, data from preferred sources or services would get special treatment; depending on the provider, that information might traverse dedicated network links, enjoy the lion's share of available bandwidth at bottlenecks, or be routed more quickly than other data. Whatever the means, preferred data would travel on the telecoms' Internet network more quickly and with greater reliability.
Telecommunications companies essentially look at this idea as innovation: they're creating new premium services which are valuable to online businesses, and which can be paid for at premium rates. An AT&T executive recently argued tiered access wouldn't change the way the Internet operates for everyday users; instead, tiered access would create new communications channels in addition to existing Internet infrastructure. Although few technical details of possible implementations are available, it's pretty easy to imagine scenarios where the telecoms' claims would be true. Many different types of businesses and organizations require reliable and secure data transfer mechanisms - think banks and financial institutions, governmental agencies, manufacturers, government contractors, and media companies. In the "old days" these industries often set up their own private networks - an example would be the networks used by now-ubiquitous ATM machines. Nowadays, the tools these industries use are based on Internet technologies, and so those businesses want to use the Internet. Through tiered access, telecoms would essentially offer these organizations the benefit of reliable, private networks with the advantages of the Internet.
But tiered access doesn't have to stop there: telecoms could partner with Internet companies like Google, Yahoo, and Microsoft to ensure their data was also handled preferentially, or let customers upgrade their service so any data they request is treated with priority - and this is where the polemics get pretty thick. Telecoms argue that charging Internet companies for preferred access is only fair, since many of those companies are currently getting a "free lunch" on the telecoms' Internet networks. Revenue from tiered access would enable telecoms to build the next generation of Internet networks.
Internet companies, conversely, argue that tiered access would create an Internet where only the richest players can get their data through reliably: unless Internet companies pay telecoms "protection money," they can't count on their data getting through to users effectively.
One objection to tiered access is technical - although, without knowing the specifics of how tiered access systems would operate, it's hard to evaluate. Generally, network engineers prefer "dumb" transparent networks to network architectures which use so-called "intelligent" data management. Historically, networks which try to move as many bits as possible as quickly and transparently as they can have achieved better performance and greater reliability than networks which analyze and apply policies to data transmission and routing. Telecoms have argued that treating the entire Internet as just one "dumb pipe" prohibits them from creating the new types of networks and data services tomorrow's online businesses require. On the other hand, the proliferation of firewalls, access technologies, security tools, and other extensions to basic Internet operations mean that, in practice, today's Internet isn't nearly as "dumb" as it used to be, although many major backbones and interconnection points are still remarkably transparent.
The Politics -- In U.S. politics, "net neutrality" is not a new topic, but it's rapidly moving from conference rooms and offices to the floor of the U.S. Congress. Several draft bills before Congressional committees in 2006 have contained provisions intended to mandate network neutrality (as of this writing, none have made it out of committee). In August 2005, the Federal Communications Commission (FCC) adopted a set of principles outlining a policy which largely embodied common ideas about network neutrality, after it had imposed a small fine on Madison River Communications in March 2005 for blocking Vonage's VoIP service to its local phone customers. Nonetheless, in legal terms the Internet is an unregulated information service within the United States.
<http://hraunfoss.fcc.gov/edocs_public/ attachmatch/DOC-260435A1.pdf>
Underlying the pro and con arguments regarding network neutrality legislation is the notion of Internet service as being subject to "common carrier" regulations developed for telephone networks (and, before that, telegraphs and railroads). Under common carrier laws, owners of a system are not allowed to leverage their ownership to undercut third-party services. In the United States, Internet service is not subject to common carrier laws. The networks which comprise the Internet are privately owned and managed, and, despite that policy statement from the FCC, they are largely free of government regulation. For the most part, network operators can do whatever they want - it just happens to have been to their advantage, thus far, to keep their networks comparatively transparent to each other.
Telecommunications companies argue that legislating network neutrality would amount to the first significant regulation of the Internet, and, in the end, would amount to government-imposed price controls for Internet services. Without pricing freedom - a concept near and dear to many sectors of the American economy - telecoms say they won't be able to roll out new advanced network services. They simply won't be able to raise the money: the bulk of revenue potential from anything telecoms can build would be gobbled up by Internet companies, many of whom would be paying little or nothing to access the network.
Internet companies and advocacy groups argue sites like Yahoo, Google, and YouTube might never have come into existence if they'd had to pay telecommunications providers for premium network access. New technologies like instant messaging, digital media, and VoIP (and, indeed, more controversial technologies like peer-to-peer file sharing) may never have been developed if tiered access schemes had been in place.
Moreover, some argue tiered access could limit consumer choice. If (for instance) RealNetworks pays for preferred access but Apple does not, suddenly RealNetworks' online music offerings can look more attractive than iTunes - not on the basis of service offerings or product merits, but simply because RealNetworks' data gets to some users faster and more reliably. Smaller businesses might not be able to get into the fast lane at all, thus preventing their products or services from reaching that critical mass to become "the next big thing."
Advocacy groups extend this concept beyond the business sector: political organizations, non-profits, and "citizen journalists" (e.g. bloggers) might have their voices squelched because they can't afford to buy into the telecoms' fast lanes. Furthermore, telecoms could simply choose not to sell preferred access to some businesses or technologies, or, more chillingly, publishers of particular political viewpoints. Perhaps Verizon thinks VoIP technology is cutting too deeply into its traditional telephone service offerings: well, there's nothing on the books which says they are required to sell preferred access to companies like Vonage or Skype. Telecommunications companies say they would never block or degrade Internet services on their networks, since doing so would be a public relations nightmare and cost them customers. That might be true, but opponents argue that U.S. consumers typically don't have enough choices for broadband service for market forces to play a major role in telecoms' access policies.
<http://www.linuxinsider.com/story/exclusives/ 50442.html>
Stay Tuned -- The debate over network neutrality will only heat up in the coming months, and you can expect punditry from both sides. In the end, it may be important to remember the debate largely originates with telecommunications companies' desire to get at a larger share of the money being earned by major players in the Internet economy, like Google, Amazon.com, Yahoo, Microsoft, and Apple, as well as whichever companies turn out to be the leaders streaming video, television, and movies to consumers. Given the economic and political power wielded by telecoms, it's almost certain they will succeed: it's just a question of how, and if the method in which they succeed turns out to be the slippery slope predicted by the naysayers.
