While not 100% 'free market,' I think that turning things like last mile lines, etc in to a public utility (allowing many ISPs to hook into them), and abolishing local government-granted monopolies is the real solution here. While abolishing the government-grants of monopoly status is unquestionably free-market, turning the last-mile lines into a public utility is slightly less so. But this move would effectively force the playing field to level a bit.
If you wanted to make it slightly less "government-y," you could just establish rules that those running the last-mile lines, and those providing the connectivity must be separate entities (and not connected to each other like some "spin it off as a separate division of Comcast but they are still parts of / owned by the same company"-type deal) so that everyone gets a fair shake.
[ That said, I'm not in favour of letting the privacy war play out in the market, because I'm not 100% sure that the market wouldn't just settle into a state where everyone was doing something I don't like to some extent. ]
It's funny that you (and many others) say this because we had this from 1996 to roughly 2004. Unbundled last element, or linesharing, was a requirement of the Telecommunications Act of 1996.
Telcos loathed it and pitched many mighty fits and threw every (physical and metaphorical) wrench they could in front of their brand new competitors. But it worked, in spite of the problems[0]. The absolute best I ever had was in Southwestern Bell territory. Over a single copper circuit I could choose from TEN different ISPs. Speakeasy, Megapath, Covad Direct, August Net, and a handful more that I can't remember. In 2002, I paid Speakeasy $160/month for 10mbps symmetrical with 8 real IP addresses and no port blocking.
President Bush was elected and the telcos' complaints were given new life at the new FCC. The unbundling requirements were swiftly removed. Now, we have this.
0 - And don't get me wrong, the problems were legion. UNE-P didn't apply to cable providers nor the new-ish fiber optic last mile buildouts. Only ILECs were subject to it so competitive providers who had built physical plants got to skate by. And there were some legitimate complaints over the "profit margin" calculation (in quotes because what's a margin in the telco business, really?). But, damn, it did work for a brief time...
> It's funny that you (and many others) say this because we had this from 1996 to roughly 2004.
Oh, I know that. The US also had more "competition" in the ISP space nationally because all of the regional telephone companies and cable tv providers hadn't yet consolidated into the mega-corps we have now.
> Telcos loathed it and pitched many mighty fits and threw every (physical and metaphorical) wrench they could in front of their brand new competitors.
Separating (e.g.) AT&T from their last mile infrastructure (and placing barriers to them re-obtaining it) would go some of the way to preventing this. Someone operating the last mile infrastructure that is not also a direct competitor to their customers[1]. It's a conflict of interest that no law or regulation is going to properly resolve.
[edit: I should clarify this. No law or regulation is going to properly resolve the situation unless we remove the incentives for the last mile operator to find loopholes to screw their customers.]
[1] the ISPs connecting to people over the last mile infrastructure
The big problem with that unbundling scheme is that it made DSL less profitable than cable internet,which didn't have to be unbundled (because they aren't ILECs).
This is why in the USA the cable companies are the big ISP leaders and the telecoms are second fiddle.
Verizon only built Fios when they got clearance that they wouldn't have to share it.
The unbundling is a great way to make sure that legacy systems are fairly priced. It's a bad way build new systems.
It's worth pointing out that Verizon has abandoned Fios expansion. Most of their build-out was to low hanging fruit like new residential communities where they laid fiber instead of copper.
So there's no simple cause+effect here related to unbundling. And even with unbundling it's not like a company loses money; they just lose monopoly rents but are guaranteed some profit whether or not they have to share.
Really the issue is about opportunity cost and financing: Verizon would rather invest in endeavors with a higher return than what they'd get with copper or even fiber. That higher return is wireless.
But there's so much cash available, or at least there has been for the past 10 years, that theoretically somebody could have stepped in to invest in these projects. Google was tentatively one of those people--initially it was enough for them to break even--but it looks like the only entity capable of committing for the long haul will be a non-profit or government entity. More so as the era of freakishly low capital costs slowly comes to an end.
> The big problem with that unbundling scheme is that it made DSL less profitable than cable internet,which didn't have to be unbundled (because they aren't ILECs).
Don't forget the fact that it's 2017, and DSL sucks. 40mbit down and 10mbit up at best is what DSL providers tend to offer. DSL just isn't viable compared to cable.
If you wanted to make it slightly less "government-y," you could just establish rules that those running the last-mile lines, and those providing the connectivity must be separate entities (and not connected to each other like some "spin it off as a separate division of Comcast but they are still parts of / owned by the same company"-type deal) so that everyone gets a fair shake.
[ That said, I'm not in favour of letting the privacy war play out in the market, because I'm not 100% sure that the market wouldn't just settle into a state where everyone was doing something I don't like to some extent. ]