How to Understand the EU-U.S. Digital Divide

Monday, October 19th, 2015

After two decades of failed efforts to spur innovation and competition in Europe’s lagging communications sector, the European Commission has promised “an ambitious overhaul” next year as part of its far-reaching Digital Single Market initiative. Until now, the EU had hoped to drive down prices for DSL-based Internet services by forcing network providers to open their facilities at discounted prices to new entrants. But that policy came at the cost of badly degraded incentives for providers to invest in new technologies, leaving the EU with minimal cable Internet, fiber-optic networks, or high-speed mobile broadband.

At least for voice services, the U.S. took a similar path in 1996, requiring local phone companies to “unbundle” their old networks at rates overseen by the Federal Communications Commission. Congress, however, wisely left the internet out of that experiment, letting fixed and mobile broadband access markets to develop largely on their own.

As a result, the U.S. has seen nearly a trillion and a half dollars in private investments for cable, mobile, fiber, and next-generation copper/fiber hybrid services. This has helped contribute to the development of innovative Internet-based businesses, where 11 of the top 15 Internet businesses, most started in the last decade, are U.S.-based, with the rest coming from China. None are from Europe.

So it’s no surprise that European regulators are now eager to reverse course and open their markets to real competition. And, conversely, no surprise that partisan critics of the light-touch U.S. regime are just as determined to make sure the EU instead doubles down on a strategy widely acknowledged to have failed utterly.

Case in point is a recent report from Stiftung Neue Verantwortung, a German think tank. The authors, an American team consisting of a legal academic and a communications professional who have long urged nationalization of U.S. internet access, admonish the EU to “avoid the mistakes” of U.S. policy and instead “continue promoting the competitive market framework that has served European consumers well.”

Even the most conservative reading of measurable outcomes of EU policy on every imaginable metric, however, says otherwise. To explore this, let’s simply compare the results of U.S. and EU policy on their five most relevant characteristics — investment, availability, competition, adoption and price. And let’s use conservative data, nearly all of it from official sources:

Investment. Most damaging to an argument in favor of continuing the European policy of mandatory unbundling is the devastating impact that policy has had on infrastructure investment. Since 1996, and even during the most recent recession, private network operators in the U.S. have spent freely, racking up investments of $1.4 trillion — over 20% of the world’s total Internet infrastructure. The result is that the U.S. now has multiple high speed networks using a variety of wired and mobile technologies, including fiber optics, high-speed cable, VDSL (high-speed fiber/copper hybrid), 4G LTE and satellite.

In sharp contrast, European network operators who are unable to recover their investments continue to fall farther behind. According to Commission data collected in a definitive 2014 study from the University of Pennsylvania’s Christopher Yoo, the wide gap in investment per household between the two economies expanded dramatically between 2007 and 2012 (the last year so far reported). In 2007, for example, U.S. network operators invested $600 per household compared to $389 in the EU. By 2012, that gap had increased by another 20%.

Availability. Despite the vast size of the U.S. and its significantly dispersed population of mountain and rural residents, access to broadband service is still nearly universal in the U.S., even if one doesn’t count high-speed mobile broadband, which the FCC so far does not. The most recent report from the FCC finds that at the end of 2013, 95% of all U.S. households had access to a 10 Mbps wired connection, with 84% having access to so-called next-generation broadband with download speeds of at least 25 Mbps. More current data from the Department of Commerce shows almost 90% have a choice of two or more wired connections and over 50% have a choice of three or more. Based on current census reports, more Americans have access to broadband Internet than indoor plumbing.

Regardless of speed or technology, broadband availability in Europe is far behind. By the end of 2013, for example, the European Commission reports that only 62% of Europeans had access to next-generation broadband services, more than 20% less than in the U.S. According to research from Aalborg University, the U.S. has deployed fiber to nearly a quarter of U.S. homes, twice as many as in Europe and a similar ratio as for cable broadband, which is almost universally available in the U.S. Outside of major cities, lightly-regulated U.S. providers already offered next-generation service to 48% of rural consumers by the end of 2013, compared to only 12% in regulated Europe.

Include mobile services, for which the FCC claims unconvincingly that it doesn’t yet have any usable data, and the difference becomes even more dramatic. The data show that 98.5% of all U.S. consumers, regardless of location, now have access to 4G LTE service, compared to only 59% of the geographically much smaller EU.

Competition. Some, like the authors of the Stiftung Neue Verantwortung report, argue that U.S. market structure has dangerously collapsed into a single dominant technology—cable broadband. And, because of early regulatory follies in local franchising rules that made cable competition illegal, most U.S. consumers face de facto monopolies of one or at most two providers for cable-based next generation Internet access.

This argument dismisses any chance of effective challenge to cable access from fiber or VDSL services from former telcos or new entrants including Google Fiber. And, they claim, the lack of competition will mean higher prices, as more consumers elect higher-speed services from fewer providers.

This idea ignores the continued importance of competitive access service below 25 Mbps, the preferred choice even among subscribers with access to higher speeds. Even assuming, as the FCC and other policymakers do, that future video, Internet of Things, and other emerging applications will create a “big bang” push to higher-speed services, it doesn’t follow that cable technology will monopolize that market. Former telephone companies, including AT&T, already offer very high-speed fiber/copper VDSL service in most of their national markets. Even now, over 60% of AT&T more than 25 million household broadband footprint can get 45 Mbps or better speeds from the company’s U-Verse service — and many can now receive 75 Mbps.

U-Verse, along with cable, has the potential for speeds approaching 1 Gigabit or faster, currently the sole realm of fiber to the home service from Verizon’s FiOS and other providers — a third source of cable competition for next-generation service that is considerably more robust than the authors are willing to admit.

Gigabit speeds are also on the horizon for mobile networks. Already, providers are experimenting with gigabit speeds in both licensed and unlicensed (Wi-Fi) networks, innovating a combination of more efficient use of spectrum, new antenna and cell design technology, and smarter smartphones that can automatically jump to less congested frequencies. The next generation of mobile service, known as 5G, will likely support speeds of up to 10 Gbps in the next decade.

Meanwhile, the competition story in Europe is bleak, despite (or perhaps the result of) mandatory unbundling and regulated pricing. The EU’s most recent Digital Agenda Scoreboard, with data through 2013, doesn’t mince words. Nearly every broadband goal and target set by the Commission has been missed. Consumers, for example, have adopted next generation speeds at a slower rate than U.S. consumers, with less than 10% willing to pay for advertised speeds of 30 Mbps and only 3% buying 100 Mbps or faster service. This despite subsidized fiber deployments in some EU capitals that the report’s authors hold up as prime examples of European Internet superiority — claims that few Europeans would find passes a smell test.

With private European network providers having little incentive to build faster networks or deploy new technologies, nearly three-fourths of EU consumers are stuck with slow DSL connections of less than 10 Mbps, often their only choice from multiple providers using the same infrastructure.

And despite strong interventionist policies, even virtual competition in Europe has failed. Incumbent phone companies, many still at least partially government-owned, must lease their networks at regulated prices to all comers, yet incumbents still control 42% of the internet market. In a third of EU countries, incumbents provide over half of the broadband subscriptions. The failure to create virtual competitors comes despite EU regulators continuing to push down the regulated price for unbundled access — the same policy that slows new investment. It’s lose-lose.
Adoption. Adoption of broadband Internet access has been impressively high over the last decade in both the U.S. and the EU, but again the more open U.S. market has delivered the better results. In the FCC’s most recent findings, which go through 2013, nearly 75% of all U.S. households had a wired broadband connection; double the world average. Already, 30% of these connections are 25 Mbps or faster. Overall, according to the OECD, the U.S. had 100 million wired broadband connections at the end of 2014, almost four times that of second-place Japan.

By comparison, the 2014 EU Digital Agenda Scorecard finds only 30 fixed broadband subscriptions per 100 people in the EU, “which corresponds to a take-up of 76% of homes. The number of subscriptions are still increasing, but the growth rate is low.” Most of those subscriptions, again, are for older DSL connections, even in markets where next generation speeds are available. (The growth rate is higher in mobile broadband, which is poised to offer next-generation speeds and, in the U.S., from multiple national competitors.)

As in the U.S., adoption in rural Europe is lower, as it is with older and poorer populations. So why doesn’t everyone have a broadband connection? The reasons are complicated, but data from the Pew Internet Research Center and others consistently show that availability and price are only secondary factors. A perceived lack of relevance and a lack of technical skills are by far the most common reasons given.

Price. In the Pew data, relevance (almost 50%) and technical issues together account for nearly 80% of the reason given for not having broadband at home. Only 10% of non-adopters cite price as their principal reason. And adoption in both the U.S. and EU is slowing as the remaining holdouts conclude, correctly or otherwise, that they neither need nor want broadband Internet at any price. Subsidizing the cost of broadband — the focus of policy initiatives in both Europe and increasingly in the U.S. — is therefore unlikely to close what remains of the digital divide.

Yet the authors of the German report ignore EU reports of slowing adoption and undisputed findings in the very surveys they cite that make clear price is not the principal obstacle for non-adopters in either market. Instead, direct data and common sense is wished away for a preferred and unsupported conclusion that it is only “America that has gotten stuck on a plateau of adoption,” a problem caused by the fact that “Americans pay higher prices than Europeans for similar high-speed Internet access products.” Neither fact is supported by ample evidence to the contrary.

In their arguments on adoption, competition, and investment, the authors of the Stiftung Neue Verantwortung report repeatedly come back to consumer pricing for Internet access as their talisman for superior outcomes under the failed EU regulatory regime. Even if Europeans have spottier access to slower Internet services they fail to take advantage of, the report’s authors conclude, EU policies can at least be credited with giving Europeans cheaper Internet than in the U.S.

Except that it can’t. Though the report acknowledges that a direct comparison of prices between U.S. and EU is “challenging,” they don’t do a particularly good job of trying to overcome the many analytic hurdles. For one thing, countries differ widely on the costs built into consumer prices for broadband. Monthly triple-play prices in both markets bundle in the cost of video content, for example, but in many European countries video content from national broadcasters is paid for separately via taxes or TV license fees imposed on each household. These fees can run several hundred Euros per year..

Broadband installation costs in Europe may also be higher than in the U.S., where consumers typically bear only modest one-time installation costs (e.g., $100). For example, while the European Commission holds up Sweden as a successful example of low monthly broadband prices for next generation fiber networks, their findings do not incorporate the standard upfront price for installation of about €2000!

In fact, an EU-commissioned study of comparative prices found that, as of 2012, U.S. prices for Internet access were lower than in Europe everywhere except at the higher speeds. And the EU study used median pricing and based its findings on advertised versus actual speeds, which likely skewed the results in favor of European providers. That’s because tests conducted by regulators in both markets concur that almost 100% of U.S. providers deliver advertised speeds or more, while in Europe, on average, “actual download speed is 76% of the advertised speed” and even worse for dominant DSL-based access.
A relatively higher price for the newest high-speed services in the U.S. (services most Europeans with access have ignored) is the thinnest peg possible on which to conclude that the EU communications policy is working while the lightly-regulated U.S. Internet market has collapsed.

And to the extent European Internet access prices have declined since 2009, as the EC finds in its latest scorecard, much of the drop can be credited to artificial reductions imposed by the regulators. But that decision also helps explains the precipitous decline in private European investment in next generation infrastructure, which in turn explains the sorry state of Europe’s Internet economy overall—the very incentive for the Digital Single Market initiative in the first place.

These measured outcomes are all based on actual data, none of it controversial, contested or convoluted. Encouraging policymakers in the EU or the U.S. to ignore the data in evaluating the obvious success of one set of policy choices and the failure of the other, as the Stiftung Neue Verantwortung report does, is simply dangerous.

The U.S. doesn’t have a perfectly functioning broadband market (or any other market), but a bi-partisan decision by Congress to leave broadband Internet access largely unregulated since 1996 has clearly worked better than the opposite approach taken by the Europeans. The European Commission is right to be working quickly to overhaul an obviously failed strategy.

All the wishful thinking and tortured logic of advocates with unrelated political agendas can’t change reality. For the sake of EU consumers, let’s hope it doesn’t needlessly delay European policymakers on the path to reform. Or provide solace to U.S. regulators who have increasingly strayed from the wise course of lawmakers.