According to a just-released report by the Information Technology & Innovation Foundation (ITIF) entitled The Whole Picture: Where America’s Broadband Networks Really Stand, “Despite the frequent claims that the United States lags in international broadband comparisons, the studies cited to support this claim are out-of-date, poorly-focused, and/or analytically deficient.” We couldn’t agree more, and extend our kudos to Richard Bennett, Luke Steward, and Rob Atkinson for a thorough and dispassionate analysis of broadband deployment and adoption across developed economies. Indeed, I suspect ITIF’s report will become the ”go to” document of the most current basic statistics on where the U.S. falls in international broadband comparisons. Given that the authors’ conclude that the U.S. presently fairs very well in international comparisons and that its position is improving, this document will not please everyone, particularly those who have long been fabricating a broadband crisis in order to support increased government control of the Internet.
While it is nice to see a well-done report on this topic, it is sad to see that such an effort is still required. Among serious broadband scholars, the broadband rankings argument is as much dated as it is silly. Given the inherent defects of such rankings, which the Phoenix Center has laid out clearly over the years (here, here, here, here, here, here, here, here, here, here, here, here, here, here, here, here), most serious scholars of broadband policy have long abandoned the discussion. Nevertheless, those more interested in revolutions than facts keep hanging on to these comparisons; though it is clear that they are finding it harder and harder to cherry pick the data to make their case. Indeed, I couldn’t help but notice that the ITIF report only listed very few names of those that continue to use the ranking argument, and those they do mention are largely at the fringe of policy relevance (a position earned, in part, by the continued reliance on discredited measures of performance and a reputation for careless research). I’d like to say the rankings debate that motivated the ITIF report was a strawman—unfortunately, I cannot.
The ITIF report covers a wide range of topics and makes the following broad conclusions: (1) the U.S. has a relatively high rate of intermodal competition in broadband markets; (2) the U.S. leads the world in 4G/LTE mobile broadband adoption; (3) the U.S. has very low entry-level broadband prices; (4) the U.S. has relatively high average broadband “speeds” with a mean of 29.6 Mbps in 2012 and 82% of American homes are passed by networks capable of 100 Mbps or better, and this percentage is growing; (5) the U.S. has the fastest deployment rate of fiber; and (6) the U.S. is not a particularly profitable market for broadband providers (see our analysis of this topic here and here). In sum, the U.S. is not falling behind in broadband by any reasonable measure of performance.
However, the authors reiterate the well-known fact that the U.S. has a relatively low adoption rate, though even so the report reveals that the country still fares well in international comparisons on this statistic. The authors conclude that “incomplete adoption will remain the Achilles heel of American broadband (p. 67).” Indeed, the adoption problem has always been the real hurdle for American broadband, a point made by many groups over time (including the Phoenix Center). The ITIF report’s authors encourage policies that focus on bringing more Americans online, but also recognize that this effort is hindered by the country’s relatively low rate of computer ownership, relatively low interest in the Internet, and relatively high poverty rate. Sadly, America has an exceedingly poor record on solving the poverty problem, and poverty probably lies at the core of our low adoption rate. A recent survey by Connected Nation attempts to quantify how much of the adoption shortfall among the poor could be remedied with subsidized rates for broadband (e.g., Lifeline and Link-Up programs). The Connected Nation survey suggests that lower rates would help but would by no means solve the adoption problem.
For those familiar with the topic of international comparisons, it is apparent that the ITIF report is the product of a substantial research effort that likely took many months to complete. Yet, as is all too common in Washington, criticism of the ITIF report by those opposing its conclusions began before the ink was even dry. In most cases, the too-quick criticism (less than 24 hours) of a serious research effort reveals little if any truth and serves mostly to expose the inherent biases or inadequacies of those doing the criticizing.
Take for example the almost immediate critique by Cyrus Farivar in ArsTechnica. (Richard Bennett, a co-author of the ITIF study and a highly competent researcher on broadband technology, has already responded (forcefully and effectively) to some comments posted on ArsTechnica. Richard is an engineer and I am an economist, so naturally my eye is attracted different material than his.) Stating the matter politely, from a trained economist’s point of view, there are a more than a few irregularities with Mr. Farivar’s analysis.
For example, the Mr. Farivar states that “the margins on [Comcast’s] Internet services … are ‘comically profitable’.” (The author admits, however, that this claim of “comically profitable” does not account for all the costs of providing service, which means his “comical” assertion is pure speculation.) The observed low overall margins for Comcast (7%) is then attributed to the fact that Comcast has “lots of other business models besides simply providing Internet access.” So let’s get this straight: Mr. Farivar’s argument is that Comcast could be “comically profitable” if it would just offer Internet access but instead deliberately chooses to have a relatively low profitability by offering other services for which it apparently makes huge losses. I suppose Mr. Farivar believes that NBC/Universal (a Comcast business) is a public service offered by Comcast at great cost to the firm and its shareholders solely for the good of the country. Personally, I’m not buying it.
But there is more. While Mr. Farivar concedes that the “U.S. has a high level of intermodal competition” between the phone and cable companies, he nonetheless laments that “there seems to be very little, if any, competition from anywhere else.” Again, let’s get Mr. Farivar’s logic straight: there are “comically large” margins in providing Internet access service, but in spite of these “comically large” margins, hardly anyone is interested in providing the service. Perhaps it’s just my formal economic training, but again I’m not buying it. (If you wish to avoid logically inconsistent arguments about competition in telecom, read this paper.) Furthermore, if Mr. Farivar is correct that no one else is interested in providing the ever-important broadband service to consumers and business, then perhaps instead of attacking the phone and cable companies a simple “thank you” to the private sector is in order both for covering over 95% of the country with broadband and for the billions of their own money they invest into their networks each year to expand and improve service.
And then we have the usual fall-back argument of the untrained: Mr. Farivar’s statement that “Americans are paying more per megabit than their counterparts in many European and Asian cities.” Let’s get this straight once and for all: price-per-megabit is a meaningless statistic. If you put a little thought to it, you’ll realize that it is not possible to measure both price and quality with a single number. (As is well recognized by those dealing with constructing a price index.) Not long ago, my broadband provider doubled by “speed” without a change in price, but I assure you I did not consider this action a 50% reduction in price. In fact, while the change in service represented a huge drop in the price-per-megabit, I didn’t even notice a change in service level. In fact, I just cut my broadband connection speed from 40 Mbps to 12 Mbps and still can’t tell a difference. My over-the-top video works just fine.
Another simple example illustrates the problem clearly. Take the prices of EBP Fiber Optics, the municipal broadband provider in Chattanooga, Tennessee. Their monthly standalone price for a 50 Mbps service is $57.99 per month, or $299.99 for a 1 Gbps service. In per-megabit terms, the prices are $1.16 for the low speed and $0.30 for the high speed service, a significant differential. By the price-per-megabit logic, we’d all be better off if the municipal provider stopped offering the 50 Mbps service and forced consumers to subscribe to the 1 Gbps service, thus paying a much lower “price” for a 1 Gbps service (with the side benefit of making us look more like Seoul). Obviously, such a decision would be a disaster for broadband in America as few budgets would permit the expense. In fact, across its services, EBP Fiber’s prices-per-megabit are $1.16, $0.70, $0.56, or $0.30. Is EPB a low-price or high-price company? The truth is that consumers don’t pay price-per-megabit for Internet service, they pay a monthly price, and their demand for service is based on the monthly fee.
Further, try to imagine a scenario where you are attempting to get a lower-income individual to pay $40 per month for a 10 Mbps broadband connection. In light of budget constraints, the individual decides that broadband doesn’t offer sufficient value to give up, say, food. Is your response to offer instead a 100 Mpbs circuit for $40, lowering the per-megabit price to a mere 10th of the initial offer? Obviously not, since you know that the decision was that the individual decided there wasn’t $40 worth of value in the broadband connection. Put simply, at any speed offered today, it’s the monthly “nut” that drives adoption, not the price-per-megabit. I suppose the individual would be very interested in a 1 Mbps connection for $4 (the same per-megabit price as the 100 Mbps circuit), but obviously it’s not financially feasible for a provider to offer a broadband connection at $4 per month. None do. Once again, this hypothetical reveals that you can’t sensibly mix prices and speed into a single index because the perversions are more frequent than are the insights.
The Chattanooga example also reveals the logical defect in the argument about government-ownership solving the broadband crisis in America because the private sector’s prices and profits are ostensibly “too high.” Presumably, the City of Chattanooga, running a government-owned and operated electric and broadband network, would never act all “corporationy” by sticking it to its constituent-consumers. In fact, EPB’s website describes its mission as follows, “As a community-owned company, our goal is not to build stock value or amass wealth. It is to help as many people in our community as possible, improving our community through reliable products and services at the lowest reasonable cost.” (Similarly, Lafayette, Louisiana’s city utility offers a 40 Mbps connection for about $50 per month (admittedly, it’s a symmetric connection, but hardly anyone needs that kind of bandwidth upstream).) These prices are very much in line with those offered by most private-sector broadband providers, and they are certainly not so low as to resolve the low-income adoption problem. Broadband is obviously expensive to provide in the U.S., and prices reflect that fact, whether set by public- or private-entities. Subsidies may help, but subsidies merely hide cost-based pricing in other charges or fees that fund them. Somebody, somewhere, must pay the full price.
Finally, Mr. Farivar extensively quotes Harold Feld, Senior Vice President at Public Knowledge and a very talented telecommunications lawyer. You may not always agree with Harold, but you’d be remiss to dismiss him. (Besides, he’s very entertaining, even to the point of having his own avatar and zippy theme song.) Harold is quoted by Mr. Farivar as asking the following question: “[D]o we want to solve our national broadband problem or not?” Harold is also quoted as saying that we need to “start dealing with the real problems right in front of our eyes.” My question for Harold is what exactly is the national broadband problem? What problem is “right in front of our eyes”? And, what are the proposed to solutions to these problems, whatever they may be? Susan Crawford wants to nationalize the broadband network. Is this a solution to the problem? Since Professor Crawford believes the government to be an industry shill, it’s not clear to me why she believes the government is a solution to anything.
In the past few years, the United States Government allocated over $7 billion to the expansion of broadband availability. And guess what? We are still looking for a measurable effect from that effort. To many, the entire effort is beginning to look like a colossal waste of money. As the New York Times recently reported, an eleven-student elementary school in Colorado has three fiber optic connections to the Internet thanks to BTOP. Similarly, the Times reported that broadband grants in Alabama and Louisiana, totaling $140 million, were terminated over undocumented expenditures and failure to adhere to construction plans and schedules. In addition, four other grants, worth $42 million, returned the money before even getting off the ground. Over half-a-billion dollars in broadband grants are now frozen. Given such waste, fraud and abuse, it’s hard to place much confidence in the U.S. government to solve any problem effectively or efficiently. In fact, that’s why we try so hard to get private-sector competition in communications markets: regulation doesn’t work very well. If it did, competition would be unnecessary. Wishful thinking, I know. (By the way, New York City requires an additional 60 million persons over its current 8.2 million to be as densely populated as Seoul, which is the highest density city in all of the OECD).
If you’re still hung up on the international comparisons debate or interact with people that are, I encourage you to read the new ITIF report. But better, I hope you find that this study (along with the other research mentioned here) has put this issue to bed once and for all so we can move on to more relevant and substantive topics in the broadband debate.