What I don't understand is why a group of people so against regulation is so willing to embrace a number calculated on the premise that an entire industry should have 1970's style utility regulations, where the government decides based on political factors how much service should cost. There is of course the natural monopoly concern. But if you want to wonder into antitrust economics, natural monopoly isn't the only market failure that warrants regulation. Antitrust concerns can arise based in network effects, the kind which led to the Microsoft monopoly, and the kind that sustain companies like Facebook and EBay today. Should we set regulated rates for what EBay can charge? (In the 1970's when telecom regulation was in force, the idea of the government setting rates for auction services of EBay's scale wouldn't be unthinkable). Or do we acknowledge that we've been trying to get rid of that sort of ham-fisted regime, and that maybe telecoms should profit from the enormous boom in demand for their product over the last two decades.

> telecoms should profit from the enormous boom in demand for their product over the last two decades.

They haven't benefited from the boom in demand as much as they have benefited from the contraction in supply[0], combined with the fact that Internet access is essentially a utility.

If it were simply a boom in demand, that could be matched by a boom in supply and customers would be better off today than they were 20 years ago. Except we're worse off - if you had broadband 15 years ago, chances are you have fewer options today in your choice of provider, you're paying more, and you're getting approximately the same amount of service[1].

[0] There was far more choice in ISP in the late 90s than there is today for most consumers.

[1] The number of people with access to broadband has increased, but that's partly due to the insanely low threshold for "broadband" speeds, as well as a function of time.

Two decades ago (1994) I (well my parents) didn't even have internet. Got DSL around 1999, screaming 256 kbps. Now, we've (well, they) got 75 mbps FIOS. We also have 3-4 providers offering 10 mbps+ LTE. People on HN like to ignore wireless providers, but for your average consumer, wireless service is much more important than wired service. The average Comcast customer uses 2-3 GB/month of data. That's easily accommodated by wireless.

> [0] There was far more choice in ISP in the late 90s than there is today for most consumers.

Kinda. They all ran over the same copper phone lines. Consumers have a lot more choice these days between cable/wireless than they did back then. Also, as capital costs go up, the number of providers in a market go down. There were a lot of companies that owned fabs in the late 1990's. Today, we're down to a handful.

> [1] The number of people with access to broadband has increased, but that's partly due to the insanely low threshold for "broadband" speeds, as well as a function of time.

No, it's because over the last 20 years telcos have invested a ton of money in their networks. Over the last 20 years, cable companies upgraded their networks from simple analog networks usable only for TV to hybrid fiber-coax networks capable of two-way data service: http://en.wikipedia.org/wiki/Hybrid_fibre-coaxial. Those upgrades didn't build themselves nor were they free.

> Now, we've (well, they) got 75 mbps FIOS. We also have 3-4 providers offering 10 mbps+ LTE.

You are by far in the minority by having access to FIOS as well as another cable provider. Most consumers only have a single broadband provider (and if they have only one, it's almost certainly not FiOS).

> for your average consumer, wireless service is much more important than wired service. The average Comcast customer uses 2-3 GB/month of data. That's easily accommodated by wireless.

First I question that statistic (it sounds unbelievably low for someone who uses Netflix even moderately). But even if I accept that, it's not meaningful to talk about a hypothetical world in which a user can rely on their mobile phone plan to provide this kind of service - essentially all users are on data plans that would be throttled heavily if they tried to rely on wireless service as their primary means of Internet connectivity.

Secondly, that's not a great statistic to use. The question isn't what people use today, given the resources they have available to them - it's what they would choose to pay for if they had access to a faster, more reliable network.

I know I would be happy to pay for proper, faster Internet access, but as it is, I'm paying through the nose for speeds that are consistently 60% of what was advertised. (This is illegal, sure, but clearly TWC doesn't care).

> Consumers have a lot more choice these days between cable/wireless than they did back then.

We have a choice between four wireless providers[0], whereas we had far more back in the early 2000s. And that only covers mobile access as far as I'm concerned (virtually no "average" user can use wireless networks as their primary access point[1]).

There is no way that you can convince me that consumers have more choice nowadays for cable than they did 15 years ago, or for mobile phone plans than they did 15 years ago. Competition has dramatically decreased in both markets.

> Those upgrades didn't build themselves nor were they free.

I'm not arguing that zero upgrades happened; I'm arguing that, for what we're paying, most consumers are not getting the level of service that they should be, and that is due to a decrease in the number of market players to a virtual monopoly, without a corresponding increase in regulation[2].

[0] Potentially three soon, if T-Mobile and Sprint merge

[1] In the US, of course. Third-world countries are a different story (look at many parts of Africa)

[2] I can't imagine Con Edison operating in NYC without the heavy regulation that is is subject to; I see no reason that ISPs should be any different.

> You are by far in the minority by having access to FIOS as well as another cable provider. Most consumers only have a single broadband provider (and if they have only one, it's almost certainly not FiOS).

The point of the anecdote is to look at how far we've come in 20 years. We're in a very connected area, but we've still come a long way in 20 years. There's lot's of places that don't have FIOS now, but those folks also didn't have DSL in the late 1990's. The average U.S. connection, according to Akamai, is 10 mbps. That's a huge increase over what it was even 10 years ago.

> The question isn't what people use today, given the resources they have available to them - it's what they would choose to pay for if they had access to a faster, more reliable network.

The prevailing trend is a massive growth in demand for wireless broadband, with limited growth on the wired side. Unsurprisingly, that's where companies are investing their capital.

> There is no way that you can convince me that consumers have more choice nowadays for cable than they did 15 years ago, or for mobile phone plans than they did 15 years ago. Competition has dramatically decreased in both markets.

Consumers don't have more choices for cable or more choices for wireless, but I think they definitely have more choices for internet accounting for all the different ways of accessing the internet. In 2000, in the northern VA area, we had 256 kbps DSL run over Bell Atlantic's wires. There were multiple ISP's, but it was the same wires and same basic service. Today, we've got FIOS, four different LTE providers, DSL, and Cox cable, not to mention satellite. Not everyone has so many choices, but I think the folks that don't have many choices today had even fewer choices back in the day.

> [2] I can't imagine Con Edison operating in NYC without the heavy regulation that is is subject to; I see no reason that ISPs should be any different.

Con Edison is in a totally different sort of business. In the last 20 years, telecom and cable companies have gone from dial-up and analog cable to DSL, digital HFC cable networks, and in some cases fiber. What's happened to Con Edison's infrastructure? Nothing besides basic maintenance. People say that telecoms should be regulated like power or water companies, but totally ignore the fact that power and water companies are using the same infrastructure they were using 50 and sometimes 100 years ago. Growth in electricity demand has been stable if not decreasing for decades. Growth in internet bandwidth demand has been exponential. Heavy regulation of the sort that's appropriate for a power company is not appropriate for an industry where we expect private companies to invest billions a year in keeping up with new technologies.

> There's lot's of places that don't have FIOS now, but those folks also didn't have DSL in the late 1990's.

That's not true at all! Most people who had cable and/or DSL in the late 1990s/early 2000s don't have FIOS today!

> The average U.S. connection, according to Akamai, is 10 mbps. That's a huge increase over what it was even 10 years ago.

Yes, Comcast has spent lots of money in expanding access to new regions, since that's the only way they can acquire new customers. They don't care about customer retention (their customers have no other choices), so they don't need to do much (if anything) to improve service to existing customers. The stagnation you're talking about has already happened.

> Not everyone has so many choices, but I think the folks that don't have many choices today had even fewer choices back in the day.

In almost all of the towns near where I grew up, the choices available today are a strict subset of the choices available in the 2000s, except at massively increased prices[0].

> Heavy regulation of the sort that's appropriate for a power company is not appropriate for an industry where we expect private companies to invest billions a year in keeping up with new technologies.

All that says is that the exact mandates of the regulation would need to differ, which I would agree is true. That doesn't mean that they shouldn't be regulated as a utility at all.

As I said before, the stagnation you're talking about has already happened.

[0] Again, I'm not counting LTE as an option, because that's not viable as a primary means to access the Internet in 2014, and is unlikely to be so anytime soon.

There is of course the natural monopoly concern.

Pointing out that it's not bad enough in facebook's case to warrant regulation does not obviate the fact that it's clearly very, very bad with the telecoms, who for some crazy-ass reason aren't classified as providing telecommunication services.

The number is way too high from what I've heard, but I'm more interested in the proposition that:

The RBOCs (now "AT&T", Verizon and Century) promised to wire up the nation with good broadband (this doesn't, to my memory, have anything to do with the Federal Telecommunications Act of 1996).

Got 10s of billions of dollars to do this.

If the two above are true, they obviously failed to do it, and weren't held to account.