Last week, NCTA, the trade association for the industry formerly known as cable, posted this amazing graph and blog post showing that the “virtuous cycle” the FCC predicted would happen when it adopted the open Internet rules (aka Net Neutrality) back in December 2010. Indeed, as the NCTA graph (based on the latest Akamai State of the Internet Report) shows, the average speed of broadband connections has not only continued to rise since the FCC first adopted net neutrality rules in 2010, but the rate of increase has accelerated since the FCC adopted the Title II Reclassification Order in February 2015. Finally, as NCTA also points out, in the approximately 10 years since the FCC first began to enforce net neutrality through the “Internet Policy Statement” and the Comcast/Bittorrent Complaint, the cost of moving bits from their source to your home has dropped 90% on a per bit basis. (Whether we are actually still paying too much because of our lack of competition in the broadband market is something of a different question.)
Perhaps unsurprisingly, this matches the findings from Free Press’ Dr. Erik Turner in this massive and meticulously documented report, “Broadband Investments And Where To Find Them.” But it’s still nice to see NCTA confirm it. One of the advantages of having blogged on net neutrality for 10 years is I can point to things like this 2006 blog post and say: “Hey, I totally predicted that. Glad to see things working as I predicted they would.” This contrasts with the net neutrality haters, who as far back as 2006 that predicted that preventing ISPs from discriminating and prioritizing traffic would result on average broadband quality getting consistently worse a bandwidth kept treating the Internet “like a truck you can just load things on” instead “of a series of tubes.”
So why did the self-appointed experts get it so wrong? And why do they still fixate on criteria like “ISP CAPEX” that neither Congress nor anyone outside the economics world cares about (and which a reviewing court utterly will not give a crap about) if better faster broadband is getting deployed as we all predicted and Congress directed?
The answer boils down to the old cliche: “Among economists, the real world is often a special case.” So while all of us out here in the real world focus on things like “hey, is broadband actually getting deployed, and is it getting better and faster and stuff so we can do all the things that make better faster broadband so critical in everyone’s lives these days,” economists poo-poo such concerns as being part of an “economics free zone.” Questioning this navel gazing in Econ Cloud Cuckoo Land will evoke sneers about how silly you must be for not understanding why the actual real world is irrelevant to the purity and wonderfulness of “real” economics. For some odd reason, a lot of folks eat this superior attitude up with a spoon and fail to ask the follow up question like “you know you didn’t actually address the substance of the argument, right?”
Anyway, I will below unpack all of this by: (a) reviewing what we actually predicted about the virtuous cycle; (b) reminding folks about the predictions of doom and gloom from the haters in Econ Cloud Cuckoo Lad (that’s a literary reference, btw, for when the usual suspects want to get all fake outragey to avoid dealing with substance); (c) reviewing why the evidence is consistent with the pro-Net Neutrality prediction and falsifies the anti-Net Neutrality prediction; and (d) why this means that if Pai tries to base his roll back of Title II/net neutrality by embracing the Singer/USTA CAPEX argument and ignoring all the other evidence, he is going down in flames in the D.C. Circuit.
(I would love include a section on what ISP CAPEX actually should look like, which casts further doubt on the question of the relevancy of any modest drop in ISP CAPEX over time as a useful measure, but I’m gonna have to save that for a later follow up.)
First A Brief Refresher — Why Do We Care About All this Any Way?
As always, the most important question in policy is “why do we care.” Generally, the universe moves on its merry way in the default pattern. I don’t feel a particular need to worry about affordable access to Hamilton tickets, much as I might like them. So why do we care about broadband and what does net neutrality, and or Title II have to do with it.
Well, we got a lot of reasons around that, you can look at a bunch of them going back to when we first got into this which include things like Democracy and stuff. But what it gets down to is that when we have broadband widely deployed and available, people incorporate it a lot into their daily lives and use it to do everything from the trivial to the life saving to the vitally important. It also turns out that as broadband gets more powerful, you can do more awesome things with it. So we keep ramping up the broadband and we keep loading stuff on it.
This is even more true today than it was back in the 00s. In the May 2017 Harvard-Harris Poll, approximately 3 in 4 voters agree with the statement “Internet access should be a right of all Americans.” So making sure that we keep getting what I like to call “biggerbetterfaster” Internet is what we are looking for here — or at least one of the things we are looking for FCC policy to promote.
And while the FCC does not have to care about my personal opinion, it does have to care about Congress’ opinion. Congress expressed the same idea about why the FCC cares about broadband in the findings Section of the Broadband Data Improvement Act of 2008, now codified at 47 U.S.C. 1301. You will note, however, Congress does not express a particular interest in how this happens or who pays for it. Congress just wants biggerbetterfaster broadband. Section 47 U.S.C. 1302, which directly tells the FCC to do that, also does not talk about, or care, who pays to make it happen or how much they pay to make it happen. It just needs to happen. If it happens on its own, cool beans. If it isn’t happening, make it happen.
Prediction that Net Neutrality Would Make This Happen v. Prediction That Net Neutrality Would Impede It Happening.
With this in mind, we look at the Net Neutrality fight between 2006-2016 and how the question of $$ for investment got so central. This broke down into two camps.
The Anti-Net Neutrality: Net Neutrality Rules Will Give Us SlowerWorse Broadband Not FasterBetter Broadband
The Anti-Net Neutrality side argued that the only way to get biggerbetterfaster broadband is to let carriers do whatever they want and, if that doesn’t work, try throwing more money at them. After all, carriers are the ones who build the infrastructure, right? So obviously we should let carriers charge the “edge providers” (Google being the big whipping boy for carriers in 2006, Netflix in 2015, and I expect it will be Amazon or someone else in 2017.) Heck, you still see carriers making this argument today.
Now a normal person would say: “Dude, providers of edge services are not ‘parasites,’ they are why I buy your damn, overpriced broadband in the first place. Did you think I had an extra $60 a month and I said ‘Ya know, I want to give this money to Comcast because I love them sooo much.’ No! I am paying you that $60/month so I can get my porn vitally important streaming video essential for my education, economic opportunity and civic engagement.” Additionally, anyone who paid attention to the “why we care section” will recall that Congress says broadband is important because of all the stuff we can do with it, not because it is a great honor and privilege to pay Comcast or whoever tons of money per month.
But you are not a resident of Econ Cloud Cuckoo Land. The residents of Econ Cloud Cuckoo Land seem to believe that people do pay Comcast money per month because they love Comcast, and that the people providing Prime Video new series The Tick! essential services are “parasites” getting a “free ride” on the carrier’s broadband connection that happens to be running to your house because you like paying Comcast $60/month.
Likewise, the anti-Net Neutrality crowd predicted that if you deprived carriers of the ability to fully exploit their investment to its theoretical maximum (because no one even in 2006 had an actual business plan for how to offer ‘prioritized’ service), then ISPs would refuse to invest even if they could make substantial profits by investing in upgrades. ”Real” economists in Econ Cloud Cuckoo Land believe that 100% rate of return on investment per sub is not sufficient incentive when carriers are prevented from potentially making even more money somehow by prioritizing. Apparently, “real” economists in Econ Cloud Cuckoo Land (ECCL) believe that not only are corporations people, but they act like spoiled middle schoolers who take their toys and go home when told they need to play by the rules.
Finally, the “real” economists of ECCL believe that corporations have a generous side. If we let them charge edge providers through prioritization as well as charging subscribers, they will use that money to invest in systems in rural areas or in minority communities they would not otherwise regard as profitable. Sure, cynical folks who believe profit maximizing companies seek to maximize profit only spend money on investments that directly contribute to profit — and that companies don’t take excess profit and use that to invest in systems they find unprofitable (or insufficiently profitable) unless forced to by regulation or as a bribe to forestall regulation. But the real economists of ECCL know better.
So, as the residents of ECCL kept reassuring us from 2006-2016, even having non-discrimination rules, let alone Title II rules, would both starve companies of funds to build in unprofitable areas and make them pout so hard they wouldn’t even build out in the profitable areas. Q.E.D.
The Pro-Net Neutrality Prediction: Net Neutrality Will Bring BetterFaster Broadband; No Net Neutrality Will Mean SlowerWorse Broadband.
Meanwhile, back in the actual real world, folks who do practical economics in the real world took a somewhat different view — including about how public policy works. Speaking only for yr humble obdn’t blogger, my operating assumptions are:
For profit companies are profit maximizing firms trying to maximize profit and minimize cost.
Where companies can make money by investing, they will do so — because they like money more than they like sulking about how regulators won’t let them play with their toys. Back when I went to school, they called this ‘rational actor theory.’ It assumed companies will generally act like rational actors, rather than like spoiled middle schoolers.
With this in mind, I ran through the following analysis back in 2006 (which I will summarize as quickly as I can). This is a drastic oversimplification, so keep that in mind when Dr. Snooty McSneeralot and the other economist in Econ Cloud Cuckoo Land start to sniff about how simplified this is.
Generally, ISPs build pipes. People sign up. Companies see potential customers and start offering them stuff to buy through these pipes. More people like this and sign up with ISPs. More companies come to sell the bigger crowd of people more stuff. More stuff is available to buy. Pipes get full, which everyone hates.
ISP says: “hey, if I build a biggerbetterpipe, I can sell the new higher speed for more $$. As long as I spend less money upgrading than the profit I make by upgrading. It’s worth doing. Because I am not only a profit maximizing firm, I am also a (wait for it) a rational actor and not a petulant teenager!” Fortunately for carriers, something called “network economics” makes it (usually) cheaper to upgrade a system than to build out the whole network originally. For example, getting cable networks to do broadband through something called DOCSIS was really expensive. But upgrading DOCSIS to get faster speeds was much cheaper than building the broadband network in the first place. Also, network operators have incentive to figure out how to build betterfasterbroadband cheaper, because $$ (especially if competition lets them pocket the profit rather than pass it on to customers — but that’s another story).
Meanwhile, the Edge Providers are also thinking: “Hey, maybe there is stuff we can do to make our stuff go faster. We could use a Content Delivery Network (CDN) to move our stuff closer to the customers so it will take less time and the traffic won’t go through those congested middle mile pipes and interconnection points. I could build my own big fiber network to distribute the content. Or maybe I could write better code and compression technology so that the stuff I’m sending actually takes fewer bits (and therefore gets there faster), or any of a bunch of other cool innovative things to get my stuff there faster so people will buy it.”
In what we call a “positive externality,” Edge Provider investing in stuff that reduces congestion doesn’t just benefit Edge Provider that made the investment. It benefits ISP (since ISP gets what the customer perceives as a better working broadband connection –customers don’t get to see who actually made the investment, they just know it works better). Even better, it benefits everyone else who doesn’t use a CDN or whatever, because there is now less congestion.
Finally, equipment manufacturers say: “Hey, there is lots of demand for our stuff, and people want more capacity so they can use the biggerbetterfaster broadband. We can make our laptops, handsets, wifi routers, whatever biggerbetterfaster so that people who buy that 25 mbps package can actually get their full value from it.
I called this “the virtuous cycle” back in 2006. The FCC called it the same thing back in 2010.
But now lets imagine the carrier has another way to make money that’s even cheaper than improving their pipes. Suppose they could get the people trying to reach the ISP’s customers to pay them for the privilege of reaching the ISP’s customers? That would let you charge twice for the same thing. That’s much cheaper than actually spending money to upgrade. Oh sure, you will eventually spend money to upgrade, but you need to make sure the “default” speed is crappy enough that people have incentive to pay to get “prioritized” speed.
Since delaying upgrades in speed in favor of prioritizing content is cheaper than investing in network upgrades, and since broadband providers are profit maximizing firms, we should expect them to do exactly that.
But if they do that (which we should expect them to do if they are rational actors), instead of getting biggerbetterfaster broadband we would get slower, crappier worse broadband for the vast majority of uses and applications. As if that weren’t bad enough, there is no no value for edge providers to try to accelerate delivery of their content to the last mile network that serves the subscriber. Whether I use a CDN to move stuff closer, or build my own fiber, or even if I drop the volume of my traffic with compression, it doesn’t help because as soon as I hit the last mile network, I get the STOP! PAY TOLL! Slowdown. And, of course, with everything moving more slowly, equipment manufacturers have less incentive to make more efficient equipment that can handle higher speeds. The whole virtuous cycle crashes and the universe becomes about as dynamic and innovative as the cable industry generally. (Which considers the high point of innovation figuring out new fees to charge you.)
So, the pro-net neutrality camp predicts that, in the real world, that we actually need non-discrimination rules to get biggerbetterfaster broadband. Allowing discrimination in the form of paid prioritization creates incentives for ISPs to slow things down that are more profitable than speeding everything up. Since ISPs are profit maximizing firms, we expect them to take this option as better for themselves, even if it is a worse outcome from a public policy perspective.
Needless to say, the “real” Economists who spend their time in ECCL instead of actual real world found this absurd and treated it with scorn. They insisted that adopting non-discrimination rules would create SlowerWorseBroadband, and only by allowing ISPs to prioritize could we get fasterbetterbroadband.
Predictions Around Title II
Then came 2015. By 2015, even ISPs were willing to say that they would build fasterbetterbroadband without prioritization (although you still got plenty of hold outs in Econ Cloud Cuckoo Land, like Chairman Pai). Now the argument centered on Title II. Real Economists in ECCL insisted that Title II would be so onerous and horrible that companies would absolutely positively hold their breath until they turned blue rather than invest. So there! Because in ECCL, corporations are not rational actors but spoiled middle schoolers.
Once again, the folks in the real world argued to the contrary for two different reasons. First, as always, there are lots of folks besides carriers who are responsible for making sure we get fasterbetterbroadband. All these companies, edge providers, equipment manufacturers, and whoever else had in an interest in fasterbetterbroadband, would have — if anything — even more willingness to invest in the ecosystem. Why? Because Title II creates a fairly high degree of certainty that the rules will remain stable. So if you invest, you have a reasonable certainty that the ISPs, who control one of the critical (and hardest to replicate) inputs for the ecosystem aren’t going to change the rules dramatically without lots of notice.
But what about the ISPs? Again, the question is how much does Title II actually raise costs and how much money do you still make. Rational real world economists looked at things like the deployment of wireless services (which took place under Title II) and said “yes, it’s true that this does limit to some degree the ability to exploit your investment because it brings certain responsibilities and oversight. But it also brings a bunch of cost-reducing benefits (like access to polls and rights of way) and lowers transaction cost to some degree. So on the whole, ISPs will still be able to make oodles and squindoodles of money by providing broadband, and the only way they can charge more and make more money is by deploying fasterbetterbroadband. So still good. (Yes, some waivers and exemptions probably necessary for the smallest players, where the uptick in cost may actually make a real difference. But this is how we’ve always done things under Title II so it should be no biggie.)
So How Did It Turn Out?
As the NCTA graph shows, and the data continue to confirm, the prediction that non-discrimination (and Title II) lead to fasterbetterbroadband is fairly clear. Oh yes, I know the difference between “prove” and “fails to falsify.” The mere fact that we have seen the market behave as we predicted it would behave is merely consistent with the theory. To actually prove that discrimination causes slowerworsebroadband, I’d need to show something more.
But what we have does work to falsify that anti-net neutrality predictions from ECCL. According to the predictions the “real” economists made in 2009, we should have slowerworsebroadband. But we don’t. We have fasterbetterbroadband. So, at a minimum, we can be confident that the anti-net neutrality prediction that non-discrimination requirements would create slowerworsebroadband are wrong. We can make a similar statement wrt Title II as well, but I would want to run Title II another couple of years before saying that with certainty.
So What’s With All This Arguing About Carrier CAPEX?
Remember how I said that among economists, the real world is a special case? Unsurprisingly, the “real” economists of ECCL did not accept the results from the actual real world. Remember, only in ECCL can you see the true reality, as revealed by theoretical mathematics as predicted by the Pythagoreans. We benighted fools in the real world see but the fleeting shadows of reality reflected on a cave wall or something. Since “real” economics predicted that non-discrimination would cause slowerworsebroadband, and that Title II would be positively catastrophic for investment, the evidence for this must show up somewhere. And since only carrier investment matters, we should look only to carrier investment. And since only investment in network hardware matters, we should exclude other forms of carrier network investment (such as developing new handsets or purchasing new spectrum). So, after diligently pruning away all investment that “real” economics said doesn’t count, Hal Singer and others found that there had been a measurable drop in carrier investment, which therefore proved that Title II was bad for investment in infrastructure, which therefore proved that Title II was eventually going to cause slowerworsebroadband.
(I like Hal, I really do. But he suffers the occupational hazard of being a D.C. economist and thus commuting regularly to ECCL. I have no doubt he feels that I, as a lawyer, am doomed to fail to perceive the true shape of the world because I’m not an economist. Fortunately for me, as I will touch on below, judges are also lawyers who are going to look to the real world.)
This proof being consistent with the rules of Econ Cloud Cuckoo Land, and being what “real” economists in ECCL fully expected to find, no one bothered to pay attention to the fact that we were actually still getting betterfasterbroadband in the real world, or that the whole point of the exercise of public policy — as made clear by Congress — is to make sure we get betterfasterbroadband. Congress and the real world don’t care about who invests in what as long as we get the results.
Enter Derek Turner and Free Press.
Derek Turner is one of those rare economists who actually lives in the real world. Unsurprisingly, the “real” economists of ECCL tend to regard him as a traitor to be disdained — rather like Slytherins hating Gryphondors as a matter of dogma. Worse, he works at Free Press, which all “real” economists despise as being all about socialism and stuff.
Derek Turner went and collected lots and lots of real world data to refute Singer’s claim that there had been any measurable drop in investment in broadband infrastructure generally or carrier investment specifically. You can read the entire 125 page report here. To summarize Turner’s arguments:
- Singer excludes investments by carriers he shouldn’t exclude, if you add these in, then carrier investment has gone up.
- Even setting that point aside, carrier investment in things that give us fasterbetterbroadband (aka, the whole point of why we care), like fiber deployment in the last mile, have gone up massively in the two years since the FCC adopted Title II.
- To the extent certain individual carriers, show lower investment in certain metrics, this is normal for CAPEX cycles and reflects predicted efficient investment rather than a pull back based on being a pouty middle schooler.
- investment is up throughout the ecosystem, exactly as predicted by the virtuous cycle.
Folks in ECCL have responded, rather predictably, by saying that if you junk all that real world stuff and use the “real economics” favored in ECCL, you see that Singer does a much better analysis than Turner, so there! This prompts much applause and beatific smiles in Econ Cloud Cuckoo Land.
Judges, Like Honey Badger, Don’t Care About Econ Cloud Cuckoo Land.
All of this finally brings us back to the question I expect readers really care about — what does all this mean for Pai’s plan to kill net neutrality? Frankly, it means that the primary pillar Pai points to as supporting roll back is going to be a big fat flop in court if he decides to go that way.
Why? You may have heard of a critter called a honey badger who, thanks to this delightful Youtube video, has the tag line: “Honey badger don’t care; he don’t give a f—.” That describes the attitude of judges when reviewing the economic evidence from Econ Cloud Cuckoo Land. Judges don’t care, they don’t give a f—.
As I noted up top, Congress actually tells the FCC what to care about, fasterbetterbroadband. Nor does Congress direct who should pay for it or how it happens. To the contrary, all Congress cares about is that it happens. If using “price caps” to make it more affordable for people to buy broadband works, 47 USC 1302 says “do that.” If “removing barriers to infrastructure investment” (by anyone, not just carriers) gets us betterfasterbroadband, then do that. Congress don’t care. Congress just wants the FCC to actually get fasterbetterbroadband deployed to all Americans. Even if we assume carrier investment is down, but we are still getting betterfasterbroadband, then Congress don’t care, and neither does a reviewing court.
Mind you, I’m not saying that you could never show a devastating loss of investment enough to make a difference. But this isn’t a question of making a prediction about something that will happen at some point in the future. It is about the nifty little graph NCTA published that shows us what we are getting now and have been getting under Title II and non-discrimination — betterfasterbroadband. And as long as we keep getting betterfasterbroadband, whether carrier investment has gone up or down is, as Derek Turner points out on page 2 of his report, rather beside the point. That doesn’t stop Derek from showing that carrier investment has not gone down (and that the decline by certain specific carriers is unrelated to Title II), and doesn’t stop Hal and others from arguing this point. But ultimately, what matters is the NCTA graph and Derek’s overall evidence and conclusion. We are getting the betterfasterbroadband Congress wants us to get.
Which means that Pai’s reliance on ECCL to reverse Title II reclassification is going to get slammed in the courts big time. I would hope that real world common sense would prevail, and Pai would back away from his ill-considered proposal. But real world common sense is scoffed at in Econ Cloud Cuckoo Land. As long as Pai continues to prefer Econ Cloud Cuckoo Land (aka The ECCL) over the actual real world, we can expect him to continue to pursue policies that don’t work in the real world and don’t pass muster in court.
Stay tuned . . .
Methinks I spy a typo of the kind that sneaks past spellcheckers:
“access to polls and rights of way”
—Alex
Does biggerbetterfaster also mean farther? As in farther out? Or farther behind? As in rural BB with slower upgrades losing out to more profitable markets accelerating buildout of bbf BB.
Don Means: One can make the argument.
Publicly owned poles and rights-of-ways are why communications using wire communications was always a Title II common carrier. The public allowed at&t, Comcast, Verizon, Time Warner Cable, etc. to put wires or fiber across and over public land. Congress established rules to ensure interstate and
international commerce could be done in an organized, fair way.
Title II is just a part of the Communications Act of 1934.
Chairman Commissioner Ajit Pai does not like the fact Title II does not clearly give the FCC jurisdiction over GOOG, MSFT, and other edge provider’s use of citizen telecommunications for privacy.
The FCC does not choose to regulate anyone and does not pick winners or losers.
The FCC follows Congressional laws and could/should now use the last sentence of Title II Section 201 to address GOOG, MSFT, and other edge provider’s use of citizen
telecommunications data delivered by an ISP.
The Commission may prescribe such rules and regulations as may be necessary in the public interest to carry out the provisions of this chapter.
Commissioner Chairman Ajit Pai should soon hear “your fired” from the Senate instead of being reappointed for five more years as an ISP infiltrator.