Note: this was a response to someone complaining about their $30/month cable bill.
While I can understand your argument in principle, I think you are overvaluing the royalties paid by the cable company to content providers as a portion of the cost to bring that content to you. For the most part the only cost to the cable company is channel integration. I would bet that maintenance of that database is nominal. Content providers make their money off of commercials, but after that, cable companies are pirates of that content. If I remember correctly, the settlement that came from those cases was that cable companies would be required to provide some number of public broadcasting channels for some number of stations they pirate. So after they have built this giant content pipe, they regulate who does and does not connect to their giant data stream in a very simple way, on or off, with very little exception. The exceptions are 1) content where per channel royalties exist (HBO, Cinemax, Encore, whatever), and 2) per program royalties channels(pay-per-view). I would expect that there is some speculation going on and the cable companies pay bulk block rates, bringing the channels cheaper to you (assuming you could even get them some other way) and likely making decent money on the side. BUT, the real business of the cable company is not the content, but the pipe. So cable companies pay for almost nothing but the initial infrastructure cost (plus the bureaucracy involved in that), then customer service, billing, and technicians and the such. One product and one price means low overhead and extremely competitive. One the cost of the infrastructure is paid off, then the money is REALLY good.
So what you pay now is a per month connection fee that for the most part is a portion of the cost to build the system that brings the content to you. Now al a carte is a request to take a very simple system and make it relatively very complicated. More equipment to control and regulate what each customer gets, these systems would of course be much more software based compared to the very dumb light switch service=on/off situation right now. The number of switches now is one per customer, based on did they pay the bill. You are proposing changing that to a number of switches equal to the number of possible customers multiplied by the number of possible channels they ever hope for the system to support (needs to be scalable). The handling of the switches would need to be related an exponentially more complicated billing system very likely bringing in security issues. Think Sigma6, in general, more things involved is always more thing to go wrong. No offense to anyone who works as a technician for a cable company, but at present it really doesn't take much of a rocket scientist to operate these networks, and even if you would disagree, you are talking about increasing the level of technical knowledge by a maintenance exponentially, meaning significantly more training, and significantly higher salaries.
So an exponentially more complicated system that personally I can only imagine would be exponentially more expensive to operate so they can more carefully micromanage their billing scheme based on something that doesn't even impact them. The only cost thing they really pay for and bill you for is infrastructure and maintenance! Why should they care at all which channels you watch? If anything, just for the sake of simplicity alone, they should just meter the time you spend watching tv per television. I think that would correlate much better than which stations you watch with regard to what costs are actually incurred by the cable company, and just embed that into the cost of the installation and you end of with a system that isn't any more expensive on the whole across the entire customer base.
Is $30 really so much? You think it would even be possible to design and implement a system where it would even be reasonable to bring you one channel for < $30/month? I would bet that an al a carte system would have a surcharge of at least $30/month before you even get any channels. The reality is that you would be paying more to get less; the necessary attraction for such a system would have to be exponential, and I would bet there are not even that many people out there that don't have cable to make offsetting the cost even feasible.
The best system to reduce the cost to the customer = ( total cost to design and build + per year cost of maintenance * number of years desired to break even) all divided by ( number of years desired to break even * anticipated number of customers ) + necessary dividends to attract the necessary number of investors. From there, once the infrastructure is paid for, the extra revenue can be used to expand into other markets. Now, dividends are going to be directly related to investor confidence which leaves customer price to be set by the price people are willing to pay that maximizes gross income, since the cost per customer is effectively negligible (this is also why they are pretty cool with letting you not pay your bill for awhile is they can keep billing you, and whatever maximizes gross income over time is their priority). This in turn tells the company the amount of time it will take for the cable company to break even. That number is directly related to risk because technology goes obsolete, and they need to cover that infrastructure cost first. So if time to break even comes out to 5 years, then likely they would say "Build it!", if it comes out to 50 years, they are likely going to say "lets do it somewhere else", cause more then likely they are investing their own money as well. Further, that time to break even will determine the exponential rate at which their market grows, and with that capital and dividends will rise over time.
Yeah, so in short, I don't think you are paying what you think you are paying, and al a carte programming is just a really bad idea for everybody.