Readers push back on Matthew C. Klein’s claim that net neutrality amounts to “one-size pricing” and “an effective subsidy” for high-bandwidth users:
I have worked for a web-hosting company for over a decade, and customers are absolutely charged for how much bandwidth they use. I’m certain that The Dish pays for how much bandwidth it uses. Klein can’t even hide behind the wiggle words “same rates” because larger customers get a volume discount and pay a lower rate per megabit than smaller customers, so in reality the all-text websites “subsidize” the video-streaming sites. For example The Dish pays more for its bandwidth than Netflix on a per-megabit basis.
The real issue here is that the ISPs are accustomed to selling more bandwidth than they are able to provide. When service ultimately degrades, they are forced to build out infrastructure to support what they have sold. Yes, this is expensive, but it’s a matter of customers getting what they paid for. One way to avoid the cost of building out infrastructure is to prioritize packets for your largest customers so they never feel the effects, while the little guy gets screwed. Even better if you charge a fee for this “service.”
The telco/cable duopoly is trying to confuse the issue. What they don’t like to pay for is Internet backbone capacity. If the traffic is on their network (local to the Verizon or Comcast network), the ISP does not pay for this traffic. They only pay for traffic that hits the “real” internet. There are many ways to reduce the internet backbone bill. Verizon/Comcast could install proxy servers that cache popular content. They could set up peering so that the Verizon network can talk directly to Amazon or Youtube. To me, this is just another cash grab, and it shows how incredibly corrupt our political system has become. The death of net neutrality will cost each of us a lot more than we think.
I used to sympathize with the argument that the ISPs paid for the infrastructure to support the Internet, and should not be forced to subsidize content providers. I see the moral and logical appeal of that point.
However, that argument assumes that Internet infrastructure is a normal commodity that can be sold or created by others to increase competition. For example, say a downtown area contains only one parking garage that only accepts Ford cars or charges $50/day. Someone else will likely build another parking garage that accepts all cars to take advantage of the irrational discrimination against non-Ford cars. Or someone will build a new garage and charge less than $50. If there are not enough cars to fill both garages, the first garage will have to lower its rate, benefiting consumers and allowing the market to work.
Internet infrastructure, like power lines or telephone wires, does not work that way. It is extremely onerous to build. So much so that only one of the most cash-flush and influential companies in the world (Google) has seriously attempted to challenge the incumbents by building new infrastructure, as it did in Kansas City. The incumbents’ infrastructure also just piggybacks on preexisting phone and cable TV networks. I concede they have to maintain the infrastructure, but it’s not like they paid billions upon billions of present dollars to build new infrastructure, the way a new entrant into the industry would. ISPs are more like utilities than a typical business in the current setup. If they want the monopoly, perhaps they should be regulated like utilities. I bet they’d love that.
Another raises a free-speech concern:
If someone can pay for faster, ‘premium’ delivery of information, why can’t they also pay to slow, or even block the service of their competitor?
For all these reasons, and having studied this some more, I’m emphatically for net neutrality – for reasons of democratic equality in online speech in an economically unequal age. Can we not have one oasis in which one argument is always just as accessible as any other. This, after all, is the great thrill of the democratic web. Every page is like every other page; that principle of a core check on power.