[Latest draft of Internet regulation bill]
Tom Vest
tvest at pch.net
Sun Nov 13 01:03:24 UTC 2005
On Nov 12, 2005, at 6:48 PM, Christopher L. Morrow wrote:
>> Are you suggesting a return to cost-based regulation? At one time
>> airline
>> prices were regulated based on air mile distance.
>
> No, I'm not, actually I think that the answer to my question was: "All
> bits cost the same to push inside 'my' network" (where 'my' is
> really any
> single entities network, and the cost is for that entity).
Is cost-based regulation so bad for critical, non-substitutable
infrastructure? That's how the US market got flat-rate Internet access.
Is flat-rate the same thing as "making some people pay for more than
they want"?
>> MCI Friends & Family charged different rates for phone calls
>> depending
>> whether the person you called was also a MCI customer. Was MCI
>> illegally
>> interfering with people calling AT&T customers by charging a
>> different
>> rate? Level 3 charges different rates for "on-net" versus "off-net"
>> traffic. Is Level 3 illegally interfering with people accessing
>> content on
>> other ISPs buy charging more? Many cell phone companies offer "free"
>> minutes when you call other people in your plan. Is Verizon
>> illegally
>> interfering with other cell phone companies by charging more? Or
>> in each
>> of this cases, are they actually charging some people less? How
>> do you
>> decide what is a "discount" or a "surcharge"?
>
> good question, I think all of the examples though have on thing in
> common:
> all the 'discount' is on 'local' traffic (local to the network),
> the cost
> differential is applied to 'non-local' traffic. This sort of goes
> to my
> point that inside a network bits all cost the same, its the external
> places that cost more... Are the folks advocating making content
> providers
> pay for 'access' to their customers willing to stand up competing
> services
> locally? (something to keep their customers who lose access to
> things they
> really care about)
There no such thing as a market price for critical, non-substitutable
infrastructure. Most markets/regulatory jurisdictions around the
world don't have / never had any cost-based pricing requirements for
network infrastructure, and in most of those markets the "market
price" of infrastructure inputs is somewhere between 90-200% of the
projected revenue potential that the infrastructure creates. That's
why there are so few operators/ASNs in most countries -- there's very
little upside for non-telcos -- and one of the reasons that markets
like the US spawned so many infrastructure users.
That may be where the Internet is headed in the absence of some major
trend shift, but is that where it should go, where we want it to go?
TV
More information about the NANOG
mailing list