ITIF Analysis of FCC Broadband Deployment Report

July 21, 2010
The FCC broadband report delivered to Congress indicates a decline in progress towards universal broadband services, yet the report has serious methodological shortcomings.

The FCC broadband report delivered to Congress indicates a decline in progress towards universal broadband services, yet the report has serious methodological shortcomings.


The FCC is charged by Congress with delivering a report each year on broadband deployment in the United States, the so-called “Broadband Deployment Report” or “706 Report” after the section of the 1996 Telecommunications Act which orders it. Congressional intent in seeking this report is clear: the FCC is to advise whether broadband networks are being deployed in a reasonable and timely manner so that Congress and the Commission can take steps to overcome any barriers to investment that might stand in the way of universal service. For the past five years, these reports have reached an affirmative conclusion, but this year’s report does not. Prior reports utilized very limited data collection methods and set a lower standard for broadband speed, 200 Kilobits/second. This year’s report employs improved data and raises the speed standard to 4 Megabits/second. The higher goal reflects more ambitious aspirations for broadband service, but has the unfortunate side-effect of excluding most wireless broadband options from the survey; most 3G broadband offerings are currently well below the 4 Mbps standard, but they’re potentially upgradeable to much higher speeds.

While broadband deployment continues to reach the homes and offices of more Americans at higher speeds, the Commission correctly concludes that universal service has not yet been achieved, and is not likely to be achieved absent a change in policy.

It’s well known, of course, that we don’t have universal broadband service in the United States, any more than any country of substantial size does, and we don’t need an annual report to remind us of that simple fact. The question that Congress means for the FCC to analyze is whether we’re progressing toward that goal in a reasonable and timely manner. Opinions can obviously differ regarding the meaning of words like “reasonable” and “timely,” so it’s useful to collect statistics year after year in a consistent manner so we can assess progress in the three dimension of broadband service that matter: adoption, deployment, and quality. But the report takes a snapshot of 2008 conditions, rather than assessing progress, to make the claim that progress is insufficient. Moreover, it fails to adequately assess the factors contributing to broadband deployment.

One reason this year’s report concluded that we are not making sufficient progress is that the FCC raised the bar on progress to 4 Mbps. While this is a reasonable speed goal for wireline networks, it is also true that there is no magic number marking a bright line between “broadband” and “narrowband” communication, so analyses of this sort are bound to be arbitrary. We should look at which service tiers are available and which are adopted. Advanced applications are limited by broadband quality, so the desire to use them provides the motivation for broadband users to subscribe to higher quality service tiers. Competitive pressures, where they exist, motivate network operators to offer higher service tiers for prices formerly associated with lower tiers. The extent to which low cost service tiers are better (faster) than they used to be is one of the important dimensions of broadband deployment that the report fails to answer.

Other countries that have adopted universal service goals have seen fit to leaven their notions of “reasonable” service with a dash of common sense economics. Australia has been cited by prominent figures in the universal broadband debate (such as former National Economic Council official Susan Crawford) as an “interesting” model. In that country, the current plan for universal broadband is to reach 93% of the population with wireline broadband at speeds up to 100 megabits/second; to serve 2% with terrestrial wireless; and to rely on satellite broadband to reach the remaining 5%. Given the high cost of bringing wireline broadband to remote areas, Australia’s plan is reasonable and sensible. It’s very difficult to justify a ten mile trench or hundreds of new telephone poles just to reach a single cattle ranch, but satellite coverage is pervasive, and with the launch of new satellites is getting better. (Satellite-based broadband has higher latency than wireline because satellite signals travel 44,000 miles from one earth station to another, but the most data-intensive broadband applications, such as high definition video streaming, are relatively insensitive to delay.)

By the logic of the FCC’s Broadband Deployment Report, Australia will never achieve universal broadband because the Australians don’t plan to deploy wired broadband universally. The FCC’s confusion of broadband service with one technology for delivering them is unfortunate.

While the report focuses on deployment, it does cite a number of statistics about adoption. In fact, the lack of adoption is by far the greatest barrier to universal broadband use in the United States. The number of Americans who choose to purchase broadband services is much lower – approximately 70% by the latest surveys – than the number to whom such services are available, well over 95%.

The Commission reports that “roughly 80 million American adults do not subscribe to broadband at home.” But most Americans in this category don’t own a computer, and therefore have no use for broadband services except as they’re required by home entertainment systems, home health monitors, or similar single-purpose information appliances. As ITIF has calculated using OECD data, if the United States had the same computer ownership rates as the average of the top 6 nations, we would rank 5th in broadband adoption, not 15th (or 12th, depending on how it is measured). Increased deployment will not close the adoption gap, but targeted adoption programs can be successful, as we’ve learned from Korea.

Subsidy programs that can support terrestrial wireless and satellite can address the increasingly tiny deployment gap, and can do so in short order. We appreciate that the FCC has chosen to use its bully pulpit to advocate universal adoption of aggressively faster broadband, but we would prefer they didn’t exclude wireless networks from their analysis. Americans have discovered the value of wireless broadband both as a “third pipe” from fixed location services such as Clearwire and as a part of the smartphone experience. Both fixed and mobile wireless networks are important elements of the broadband experience today, and the infrastructure that supports 3G broadband can be upgraded to 4G easier than completely new networks can be deployed.

As to the question of whether we are making reasonable progress, the Commission’s write up makes no reference to trends. A robust definition of progress would present longitudinal data to show trend lines of progress toward deployment. And only when the trend lines become flat would the claim that progress has stalled be justified. Yet Chart 1 of the Appendix to the report (the “High-Speed Services for Internet Access: Status as of December 31, 2008” report released Feb. 2010) shows continued progress in fixed broadband. In fact, according to their data the rate of growth accelerated in the most recent half year, with new lines growing at a 30 percent faster rate than before. This does not suggest a lack of progress.

The report does rely on the national Broadband Plan findings to correctly argue that households in areas that are costly to serve are unlikely to get wired broadband absent subsidies. But the report provides no evidence for its assertion that households with low speeds are unlikely to ever see higher speeds. Yet, the report acknowledges that speeds have increased consistently, citing the Broadband Plan which states that “the average advertised speed purchased by broadband users has grown approximately 20% each year for the last decade.” There is no evidence that speeds will not continue to rise. This points to a problem in defining counties as unserved if they don’t have access to 4 Mbps service. A county with 2 Mbps service or 3G service is very different than one with no service at all, for the simple reason that it is much easier to upgrade broadband speeds than it is to deploy entirely new networks. Moreover, the Commission states that broadband – defined as 4mbs – is not being deployed to all Americans in a reasonable and timely fashion. Yet as the National Broadband Plan concluded, getting to 100 percent wired deployment would be require extremely large public subsidies.

Therefore, for the next 706 report the FCC should focus more on trends and on better discerning households (and counties) with access to some broadband, even if it’s less than ideal, and on the reasons for the deployment rates.

Finally, it is clear that absent more aggressive broadband policies we will not make the kind of progress we need to make as a nation; ITIF has argued this for years, and National Broadband Plan reached the same conclusion. The key question is what kind of broadband policies will be most effective at stimulating both deployment and adoption. Unfortunately, the FCC report is quite vague when it comes to this critical question. If we are not making the progress we need because of flaws in the marketplace and insufficient efforts by broadband providers, that is one thing; but if it is because it’s just too expensive to deploy broadband in certain areas without subsidies or that people don’t have the requisite digital literacy levels to use broadband, that is quite another thing.

The report alludes to this by citing the statute which states that if the Commission finds that broadband is not being deployed in a reasonable and timely manner, it must “take immediate action to accelerate deployment of such capability by removing barriers to infrastructure investment and by promoting competition in the telecommunications market.” The report goes on to state that the FCC has already begun to take steps “by removing barriers to infrastructure investment and by promoting competition.”

But it’s not always barriers and lack of competition that are the problem. In fact, the 1996 Telecommunications Act was written 15 years ago, long before we understood how to correctly diagnose the deployment problem. Barriers to infrastructure investment exist in some places, as the national plan notes, but they aren’t the major barrier to deployment. There is no evidence that the lack of competition is deterring investment in high cost areas. In these situations, the cause of non-deployment is pretty simple: it costs providers more to deploy and provide broadband services than they can earn in revenues.

That’s because: 1) densities are low and required infrastructure investments high; and 2) likely take up rates are low, because these populations are generally lower income and have lower levels of digital literacy. These problems are not to be solved by regulation or “more competition,” they must be addressed by subsidies for both deployment and adoption. On this front, as ITIF has shown, the U.S. has simply failed in comparison to the world’s broadband leaders. And while the stimulus legislation helped remedy this, it did not go nearly far enough. Unless the FCC, the Administration, and Congress step up the plate and provide real funding to close the broadband gap, there is not much point in issuing 706 reports in the future. The reports are really an indictment of government’s failure to provide active policies to spur adoption and deployment, especially broadband subsidies.

The bottom line is that the universal service fund for traditional telephony needs to be replaced with a universal broadband fund that brings cost-effective service to remote areas, and federal funding is needed for programs to stimulate more widespread ownership and use of computers. The debate over the Broadband Deployment Report’s new methodology and curious emphasis should not obscure this fact.