Network television advertising accounts for approximately $20 billion in revenues for the networks. This accounts for approximately 100 million households, so Kellner's estimate of $250/year per household is certainly reasonable.
That $250/year may sound like a lot to pay for broadcast television, but when you consider the hidden costs of the current system, you might change your mind.
First of all, those $20 billion/year in advertising fees are ultimately paid by you and me. Every time you buy a product which is advertised on the television you pay a premium. It may be a few cents here and a few cents there, but even if advertising did not increase corporate profits the average household would still spend around $200/year extra (98% of households own televisions).
As if it wasn't enough that the television companies are claiming to provide a free service which you the consumer are ultimately paying for, they're doing it over the airwaves owned by that same public. One thing I didn't realize until doing research for this article is that television broadcasters receive the rights to the public airwaves absolutely free. So on top of the real cost, in terms of increased product prices, we the public are paying the opportunity cost of not being able to use the airwaves for a better purpose. When you think of that it's absolutely mindboggling that Kellner has the nerve to say that commercial skippers are the ones doing the "stealing".
Putting these issues of implementation aside, I think it is important to have some broadcast television freely available to the public. News and education dissemination are certainly goals which most of the public agrees is a worthy cause to allow the use of the public airwaves. The line between entertainment, news, and education is rather thin, and I don't think any legislation should try to draw it more clearly. But these functions need to be served with the public's interest kept as the goal, and not by corporations driven by profits.
I think the first step is to take all of the current television spectrum and allocate it to 2 or 3 channels given to non-profit charity institutions. If the demand from these institutions is higher, then more space could be allocated, but other than PBS and perhaps one or two other local stations I doubt there will be very much demand.
You might wonder why I don't favor direct government production of network television. The reason for this is I believe there would be too much danger of the station being used for government propaganda. Filtering the money through tax breaks and then into voluntary contribution or other transactions with non-profit charities seems to me to be a less dangerous alternative.
The rest of the spectrum I would allocate to three competing non-profit charity institutions which will implement IP broadcasting. Each of these institutions would then reallocate the bandwidth according to various methods. Some could be used for non-profit news, education, and entertainment dissemination. Some could be auctioned to one-way high speed wireless ISPs. Some could be re-auctioned on a packet-sized (and time-of-day based) basis directly to broadcasters who would need to encapsulate whatever content they are broadcasting into IP. For instance, encrypted pay-per-view television over IP would be one possible usage. After a year's test run, lease rates would be set to provide the government with revenues for the use of the public airwaves.
Admittedly, my plan is not completely worked out, but I feel it is much better than giving the public airwaves to billion dollar corporations to use for spreading propaganda and profiting off of it.