Is a "neutral" net anticompetitive?
March 22, 2007
The "net neutrality" debate is a complicated one (witness Google's recent twists and turns). Take the very important issue of competition. On the surface, it would seem that those in favor of making net neutrality the law of the land are fighting the good pro-competition fight. By preventing telcos, cable operators, and other pipe owners from giving favorable treatment to certain forms of data - allowing, say, video from TV studios to flow faster than video from amateurs - a net-neutrality law would keep the playing field level for the little guys.
In theory, that's true. In reality, it's a little more complicated.
Net neutrality exists in the abstract, in the realm of protocol. Because the content of any packet of data is invisible to the pipe carrying it, by protocological fiat, every packet is treated the same. If that was all there was to it - if theory and reality were one - then pro-neutrality would mean pro-competition. But it's not all there is to it. In addition to the abstract realm of protocol, there's the very real - very physical - realm of infrastructure. Regardless of protocol, superior infrastructure provides superior quality of service - ie, faster, more reliable transmission of data. To put it a different way, a company can buy a competitive advantage by buying (or renting) better infrastructure. So, for instance, if I have the money to contract with a caching company like Akamai to speed the delivery of my content, then I have an advantage over the saps who can't afford such services.
As Akamai itself puts it: "Akamai's technology ... has transformed the chaos of the Internet into a predictable, scalable, and secure platform for business and entertainment. The Akamai EdgePlatform comprises 20,000 servers deployed in 71 countries that continually monitor the Internet – traffic, trouble spots and overall conditions. We use that information to intelligently optimize routes and replicate content for faster, more reliable delivery." No wonder so many companies use services like Akamai's - who wants to be stuck with all the little guys in the "chaos"?
Protocol is neutral. Infrastructure isn't.
If net neutrality becomes law, it would prevent big companies from locking in an advantage at the protocological level - giving certain types of data privileged status - but it would allow big companies to lock in an advantage at the infrastructural level. And who has the best infrastructure? Well, Google, of course. Through billions of dollars in capital investments, it has created a kind of shadow internet for the express purpose of providing its content and services with an advantage in transmission speed and reliability. That's great, because it means we all get our search results that much quicker. If the net were truly neutral, truly agnostic about what it carried, we'd spend a lot more time twiddling our thumbs.
But here's the downside. As Google shifts into content and services businesses, that very expensive and very sophisticated infrastructure turns into a big entry barrier for would-be competitors. Think of the software-as-a-service (SaaS) market, for instance. For SaaS providers looking to serve businesses, the speed and reliability with which their applications run through the browser window are absolutely crucial to success. If I'm a small startup looking to compete against a Google (or a Microsoft or any other large company able to invest many hundreds of millions of dollars into its network), I start out at a big disadvantage at the infrastructural level. There's no way in hell I can afford to build the kind of infrastructure that the big guys have. But perhaps I could, at a far lower cost, contract with the pipe owners to give my service privileged status. In this scenario, dismantling "net neutrality" (as commonly defined) could actually be pro-competitive by helping to counter the infrastructural advantages held by large companies. Embedding "net neutrality" into law would, by contrast, strengthen infrastructural advantages, creating ever larger barriers to entry over the long run.
I'm not trying to argue that protocological neutrality is a bad thing, and I'm certainly not suggesting that pipe owners should be trusted to promote competition. I'm just pointing out that it's a dicey issue. Over the long haul, which would turn out to be more anti-competitive: a Net rendered non-neutral by protocol, or a Net rendered non-neutral by infrastructure? I don't know. It's a very good question to debate. But let the debate begin with an honest admission: The Internet is not neutral and never will be.
As for those who would look to politicians and lobbysists to maintain the net in its putatively Edenic state: Be careful what you wish for.
"There's no way in hell I can afford to build the kind of infrastructure that the big guys have."
But small and growing companies have many ways to steadily scale up their capacity as they grow, especially by outsourcing their hosting. You mentioned Akamai, which rents capacity to companies, and there are many levels of managed hosting that aren't especially expensive or require any large up-front capital expenditure.
And this is theoretically true for telcos charging different prices for different tiers of service: as long as any customer can buy higher-tier service at a price proportional to their needs, the cost ought to scale up as a company grows and can afford the higher tier. Of course there are many reasons to distrust the telcos and their motives in this area -- a lot of it looks like a blatant attempt to soak Google, et al. How exactly tiered service would be priced is a big question.
The Economist has covered some of these ambiguities in the net neutrality debate pretty well, in contrast to some of the good-vs.-evil coverage and commentary elsewhere. One of the unfortunate things about debates like this on the web is that they have an even more relentless political correctness than debates in the "real world."
For better or for worse, it reminds me of this one:
If health care is provided to everyone for free, rich people will still have their better paid services; and as they do that, the services for poor people will languish because the rich people no longer argue for better free services . . .
I agree with Kendall - you can just rent these services. The same way you'd rent a phone line. For most applications, getting a fast response time to the user is more about cutting images and other flab from pages, not big-iron infrastructure. I've seen pages with 50+ object [inc. 150kb of Flash] stuck into them - to dialup users they might as well not exist.
The big issue with infrastructure is BitTorrent/Skype/etc soaking shared consumer connections. Go to an open WiFi node in a coffee shop - if your connection goes sloooww, they'll be someone else who's just started talking to their laptop. Traffic shaping is here to stay. People need to start paying for what they use. Sad, but it's got to happen.
There's a big difference between discriminating Skype traffic from HTTP traffic, and, discriminating Democrat traffic from GOP traffic.
Another point is that, as a startup, you don't *have* to build your application as a traffic-heavy web application. There are a number of old and new tehcnologies (look at http://www.visualthesaurus.com) that allow you to develop rich applications that minimize the amount of data that needs to be sent between your servers and the user's desktop.
Post a comment
Thanks for signing in, . Now you can comment. (sign out)(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)
"Riveting" -San Francisco Chronicle
"Rewarding" -Financial Times
"Ominously prescient" -Kirkus Reviews
"Riveting stuff" -New York Post