Here in Sunny Alberta rough estimates for the dollar cost of setting up a new ISP with satellite access to a major US connection point come in units of a hundred thousand - but that only covers facilities, software, people, and initial marketing. To actually sell the service, I'd need regulatory approval - and cost of getting that starts in the millions.
When you're running a company with a few billion a year in revenues, spending a couple of million on regulatory compliance is a triviality, but because it's likely to be a serious multiple of the first year's expected revenue for a local start-up, that start-up is prevented.
One result is of this is that the big players aren't price constrained by competition - and so set pricing to make a positive margin on the cost of regulation.
Another is that the competition they do face comes from only two or three other big players all of which play by the same rules, use the same technology, hire the same people, and avoid disruptive innovation.
And, of course, the real hutzpah zinger in all this is that the same consumer who pays the cost of regulation directly to the ISP and indirectly through being locked out of technology change and opportunity, also pays for the regulator through the tax system.
Put this stuff together and you have the reason the biggest companies in major markets generally tend to favor heavy handed regulation - and therefore why Google, among others, has spent tens of millions pushing for "net neutrality" regulations in the United States.
(Net neutrality is not, of course, about neutrality - it's about having government control and monitor what carriers are allowed to transmit, to whom, and at what rates with specific and immediate benefits to bandwidth hogs like youtube and specific and immediate limitations on premium services contracts like those Apple put in place with AT&T to give their iPhone a performance advantage.)
Net neutrality was set back a bit two weeks ago when the U.S. Court of Appeals for the District of Columbia ruled that the FCC does not have the statutory authority to impose traffic management on Comcast. Unfortunately, however, the bureaucracy now driving this has "forever funding," courtesy of the U.S. Treasury and isn't going to stop until it succeeds - either by getting a favorable court decision or through a mandate change enacted by Congress.
The betting, therefore, isn't on whether or not the regulators will get their way, it's on whether enough conservatives get elected in November to delay this long enough for it not to matter - and that's where you come in (assuming you're American, of course), because there are things Americans can do.
First, you need to learn enough about net neutrality to understand what it really means, who's pushing it, and what its likely consequences will be - just read some of the stuff being written by proponents of the scheme and keep asking yourself three questions as you go along:
For most of us knowledge is power - so learning more about it should give you both the means and the incentives to warn others, but if you've got appropriate network management responsibilities there may be something else you can do to affect this, particularly if you think the ideas being used to sell net neutrality sound attractive.
It's this: just go ahead and net neutrality on your own network and for your own users. Day one you're going to find that net neutrality requires you to give incoming porn packets exactly the same forwarding priority on your network as text messages to Sales or voice traffic for the CEO's office - and as soon as you decide to block one set while giving the other a priority boost, you'll have both demonstrated the fundementally Orwellian nature of the whole net neutrality sales pitch and turned yourself into one of its opponents.