Not News: Deploying Broadband Costs Less in Dense Cities

Riga, Latvia

An unfortunate idea continues to circulate in the mainstream press, fueled by broadband populists: in this case, a piece recently published by The New York Times proclaims that the “U.S. Struggles to Keep Pace in Delivering Broadband Service.” Such headlines perpetuate an argument that, although appealing in its simplicity, is ill-founded and dangerous in its policy implications. The argument goes something like this: broadband is a utility, a utility that the U.S. is “falling dangerously behind” in providing; therefore governments should step in, either through regulation or public provision of infrastructure to ensure we have top-notch utility service. Forget the politics of such a position – the premise of the argument is simply not borne out by the facts. For starters, our past research has shown that by many measures the U.S. ranks competitively in international broadband speeds, prices, and deployment. But let’s look at the specifics laid out in the Times piece.

The author, Edward Wyatt, attempts to draw policy conclusions by comparing the broadband speeds of San Antonio and Riga, the capital of Latvia. As a threshold matter, it is not clear that we can conclusively say that Latvia offers better broadband than America. While Riga boasts impressive speeds, the FCC’s International Broadband Data Report shows the U.S. leading Latvia in the percentage of households with broadband access. Furthermore, Akamai’s latest State of the Internet Report, which is generally more accurate than the self-initiated tests of Ookla, measures Latvia’s average speed a mere 1.9 Mbps faster than the U.S. – far from an alarming discrepancy warranting policy change.

In fact, it would be surprising if Riga did not have broadband speeds higher than San Antonio. Low population density poses one of the main impediments to building out high-speed networks in the U.S.  The low population density of America’s metropolitan areas (we are largely a suburban, single family household nation) compared to other countries make U.S. broadband networks significantly more expensive to build per-subscriber. The initial costs can vary tremendously depending on the density of a population served by a network. The U.S. population is spread out – even in cities, the U.S. tendency towards suburban sprawl make broadband networks more expensive to build. Wyatt’s comparison is a prime example: Riga has roughly twice the population density of San Antonio in terms of persons per square km – it would be much more costly to build a high-speed connection to a given number of customers in San Antonio than in Riga.

Moreover, the United States differs from many other nations in that major U.S. broadband providers generally do not “cherry pick” densely populated urban areas to deploy to. Therefore, they have to charge the same price whether they serve low-cost, densely populated neighborhoods in cities or high-cost, less densely populated areas at the edge of metro areas.   In contrast, a 2010 Latvian economic report admits that the country’s high average broadband speeds are due to fiber “mainly located in Riga and the largest towns in Latvia.” These dense cities supported fiber build-out by private investors – indeed, Riga’s networks are not municipalized, as might be expected given Wyatt’s line of thinking. The reason San Antonio has slower download speeds than Riga isn’t due to some default in our network, fixable by government intervention; this is a simple problem of lower demand not yet justifying the cost of ultra-high speeds deployed to single family houses.

As would be expected, Latvian Internet providers have trouble serving its rural populations. The EU has recognized Latvia’s large discrepancy between rural and urban broadband access as a problem – a problem that justified the EU investing the equivalent of $139 million to improve rural broadband infrastructure. Adjusting to a per-GDP basis, this would be like Congress allocating $82 billion to rural broadband infrastructure – 11 times more than the BTOP grants made as part of the stimulus package. In other words, superior Latvian networks don’t come from more competition, more enlightened providers or other factors.  They come from more densely populated cities and massive government subsidies.

The Times piece quickly touches on another important reason Riga sees higher speeds than San Antonio: Riga had no Internet infrastructure in place before private investors began delivering fiber to the premises (FTTP). American Internet providers have invested large amounts of capital in legacy cable and telephone (DSL) networks that largely serve the demands of current broadband applications. As Mr. Furchtgott-Roth points out in the article, we should stop thinking of the provision of broadband as a race. Simply because some cities with sufficient density were able to leap-frog over the U.S. by building networks with brand new technology does not mean that we are permanently “losers.”

As demand for higher broadband speeds increase, economics and consumer demand will support more build-out. We are already seeing fiber networks being deployed in some cities and organizations devoted to rooting out areas where demand would support the high cost of fiber. Furthermore, cable operators have been investing in upgrading to DOCSIS 3.0 (Latvia, by the way, has virtually no 3.0 build-out). We are also seeing the potential for real competition from wireless LTE in the broadband space – a technology that the U.S. leads the world in build-out, and which Latvia lags woefully behind.

We have to stop simplifying international comparisons of complex, technology-based industries to a single number. We certainly agree with Wyatt that broadband is important to economic growth, but the simple fact that some cities have faster Internet than others does little to guide policy. For our nation’s cities, intermodal competition will continue to serve us well. For many of the less densely populated rural areas, carefully targeted government grants like those in Latvia make sense. But the notion that broadband speeds in select foreign cities, where private investors were able to build advanced networks to dense populations at low cost, can justify a wholesale shift in our competition policy just doesn’t add up.

Photo credit: Flickr user Ullisan

Print Friendly

About the author

Doug Brake is a Telecommunications Policy Analyst with the Information Technology and Innovation Foundation. He specializes in broadband policy, wireless enforcement, and spectrum sharing mechanisms. He previously served as a research assistant at the Silicon Flatirons Center at the University of Colorado, where he sought to improve policy surrounding wireless enforcement, interference limits and gigabit network deployment. Doug holds a law degree from the University of Colorado Law School and a Bachelor’s in English Literature and Philosophy from Macalester College.
  • Rui Castro

    Agree with your comment that as a general rule we must not compare apples to oranges.

    Being a EU citizen I must also stress that, while I am all in for broadband, I am against the governments subsidizing fiber build-outs by giving away public monies to incumbents and established telecommunications providers. This applies to the US as well.

    This brings no added value to the economy as a whole as most of these invested monies end up going outside to equipment and fiber manufacturers. The only added value is digging up trenches and running cables (does not apply so linearly to the US).

    Working for a telecom operator I also see no value in the two methods chosen to subsidize:

    1) extend credit line/give away money to the established operators – this only results in everybody building out the same areas with multiple unused fiber runs while not covering every household which, by the way, would be much cheaper (government/regulator lack of insight/vision);

    2) promoting public competition for build-out of less commercially interesting /geographically challenging locations, while not mandating service obligations (this in turn gets the winning bidder to generally inflate their investments so as to pay for the build-out with public funding, while providing an illusion that there is a competitive wholesale offer that in reality no one will take, since it costs more to lease than to build a new one)

    Sorry for the rant. To end, correct that slip of the pen there, where you stated that cable is legacy technology. When everybody was talking FTTP/B/H, cable was doing FTTN and was already what became known as a Next Generation Network. The 3.1 evolution for (Euro)DOCSIS will take it even further.