You also have to remember that that markup has to cover overall plant installation costs. You can talk about amortization all day, but if you want to be fair you should break out initial installation costs as a component of HSI service, since I know plenty of HSI-only Comcast subscribers that would shift the “plan cost bias” toward HSI, if you follow. So (being liberal, yes, but for the purposes of argument) if we set an installation cost for cable at $1000 (a reasonable number, considering when the plant was installed) and do a 25% on that, we get $250. Spread this over ten years and you get another few bucks per month, or another five percent added to the cost vs. revenue equation.

Creative accounting? Maybe, but these calculations bring to light why everyone isn’t going out and building networks left and right: it’s a long-term deal. If rabid investors (the only type there are these days) wanted to do calculations based on three years instead of ten, on the above network costs you’d be looking at 250 / 36 = $7 per month in amortized infrastructure costs for internet, with the 25% x $1000 number. That’s 17% of a $42 average HSI bill right there; add in your other 35% and you get over half of revenues as costs. So the profit margin is now below 100% and investors in a new network also have to hem and haw about competition from cable and DBS. So the next-gen network never gets built. Sad, but that’s apparently how things work.

That said, Comcast isn’t paying anyone for most of its traffic nowadays. It’s all peered of. What goes over Level3 is probably billed at 95th, so it’s in Comcast’s interest to keep working out peering deals to minimize this traffic, but that’s about it. I”ll bet Comcast’s entire bandwidth bill for transit (vs. peering costs, which I’m sure are pretty high) pale in comparison to what Google pays for its bandwidth…and Google has a backbone and peering of its own!

There IS NO fair form of consumption-based charging for bandwidth! Unlike processed water, gas, and electricity, bandwidth is not generated by backbone providers or ISPs, nor is it something consumed in the process of transmitting it. With the fiber and copper infrastructure paid for long ago, there is no expense other than maintaining the existing equipment – which hardly varies, whether the networks are idle, or fully saturated. The small difference is mostly due to the likelihood of more service incidents when there are more individual customers on a given network. So many people simply do not understand this, because they’re trained by traditional utilities to think of their bills in terms of what they really have consumed.

The only fair form of bandwidth billing should be by provisioned_speed x billing_period. And if the ISP is not capable or willing to reliably provide the provisioned speed, customers can downgrade to a cheaper tier and thus avoid unjustly enriching their provider.

These networks should have been upgraded long ago to reflect the needs of the customers, and the times, but instead the telcos choose to keep their profits and publicize a false “shortage” to provide them with a pretense for raising prices. It’s absolutely outrageous that they try to play the victim here. This is an EMBARGO.

There is NO SUCH THING as a “bandwidth hog”. Customers with “unlimited bandwidth” service are provisioned with an “up to” maximum speed throughout the course of the 1-month billing period. Noone has ever actually obtained the full bandwidth possible to them. Additionally, the ISPs place too many customers on each segment to carry that much bandwidth to each of them, but that’s not the fault of the customers.

Note that I’m not arguing that ISPs don’t get *billed* by their upstream providers for bandwidth. Usually, the bill is static, and provides for pretty much “unlimited consumption” as long as upload/download ratios remain even, and the downstream ISP doesn’t frequently use much more than their provider has reserved for them. However, these bandwidth rates are the result of a cartel, and not an open market.

Small ISPs aren’t talking about the same kind of bandwidth expenses as top-tier backhaul providers (and major ISPs who are their own fiber/backhaul moguls). Small ISPs are collaborators with the cap-and-overbill camp because they have no choice. Top-tier backhaul providers do not suffer any sort of expense for each gigabyte they convey, although they do keep track for billing purposes. Like the small ISPs, they make money via “billable events”. Their profits are not tied to the quality NOR quantity of service they provide. Their profits are simply tied to the arbitrary terms of their contracts. If the big ISPs and telcos are raising their rates, then the little ISPs have an opportunity to “regretfully” send out emails announcing a raise in rates. “Business realities” and all that jazz.

It’s impossible for telcos to establish why they deserve to be paid according to what their network is used for. An analogy of “consumption-based billing” for bandwidth is where a telco charges phone rates according to how many words are spoken or received by the parties on the line. They don’t care how long the call is, just how many words. And, just to be “fair”, they have a higher rate for customers who speak rapid languages such as Spanish. To do this, they have agents who monitor the lines to determine what sort of conversations are taking place. Did you conduct a profitable transaction over the phone? Fork over a percentage to the telco; we can’t have you cheating them, right?

Answer to your “up”: Because the “unofficial” Net Neutrality rules that the FCC has been operating would prevent that (just ask Comcast). But the telecon/cable lobyists are trying to change that and prohibit any effective Net Neutrality rules.

Regarding “Why would I want to watch TV on my computer in the first place.” The latest crop of TVs now have Internet connectivity – so you can watch TV from the Internet (plus there are now lots of STBs & Blu-ray players with the same connectivity). Seeing this growth of technology is EXACTLY what scares the cable side of the business.