Hacker News new | ask | show | jobs
by hellojesus 1164 days ago
Radio frequencies are not public goods. Public goods must be nonrivalous and nonexcludable. But only one entity can effectively transmit on a given frequency at a time, so it is definitely rivalrous.

Seems the argument should be more that it is a merit good, but even then some determination must be made as to how to allocate each frequency resource. Auctions are a decent way to do that; supply and demand. So long as they are public and truly go to the highest bidder, I'm not sure there is much to be dissatisfied about.

Thankfully the internet exists to allow people to transmit whatever they want; granted it's still limited to the internet so not the same.

2 comments

The idea of "only one transmitter" might have made sense in 1950 for analog signals with no chance of replay, but I have to think we've learned a lot about multiplexing and collision detection.

If we allocated broadcast bands today, would we have a hundred narrow channels, or would we have a smaller number of channels and use a TDMA or CDMA style system to expand the number of available slots? I could even imagine packetized broadcasts-- rather than "KXYZ-FM" being on 98.6, your receiver would just scan the entire band for packets tagged KXYZ, and they could dynamically reallocate based on local noise levels or atmospheric conditions.

That sounds a bit like DAB multiplexes. Except that there is a single transmitter and you buy a slot in it. Cognitive radio with dynamic frequency selection was all the hype along with TV white space radio. But the only actual service that has materialized was the 3.65GHz CBRS one. And that seems to only have been successful thanks to it sharing frequencies with the 3.5GHz 5G bands, making equipment available and affordable.
>>Public goods must be nonrivalous and nonexcludable.

Says who?

>> Auctions are a decent way to do that;

Clearly they are not, nor it is how that is done, as entire segments are portioned off for a use, then auctioned off for that use.

>So long as they are public and truly go to the highest bidder, I'm not sure there is much to be dissatisfied about.

Lots. for one the fact that companies can buy up huge parts of the spectrum for "future use" and sit on it for decades denying access for innovation.