Aujourd’hui ce qui ne vaut pas la peine d’être dit, on le chante.
This line, from Le Barbier de Séville, is translated as, “Nowadays what isn’t worth saying is sung.” International comparisons of broadband services certainly fall into this category, and this week the New America Foundation is singing again with a 2013 update to its 2012 Cost of Connectivity Report. While New America’s 2013 Report has garnered some glowing accolades in the press (see, e.g., here and here), the hard reality is that New America’s 2013 Report continues to commit all of the numerous technical errors I highlighted in my earlier blog critiquing New America’s 2012 Report. Thus, as New America has decided to press ahead without correcting its earlier fundamental analytical flaws, some additional thoughts are warranted.
At bottom, New America’s claims that “U.S. cities lag behind cities around the world” and that “American consumers take high prices and slow speeds to be a given” are based upon a comparison of the “prices” for services across cities both domestic and international. Suspending disbelief on the price calculations for the moment (again, see my earlier blog), the next logical question to ask is what reason do the authors give for this “laggard” status? The 2013 Report concludes that this laggard status is the direct result of a lack of competition, claiming that their evidence “shows that the most affordable and fast connections are available in markets where consumers can choose between at least three competitive service providers.” Despite this claim, however, New America presents no systematic evidence on the number of competitors (see my earlier blog and Brent Skorup’s blog for similar and other valid criticisms), but, again, let’s suspect disbelief for the moment and go with the flow. Because New America’s analysis concludes that our purported laggard status is somehow the result of insufficient competition, New America offers up but a single recommendation: i.e., policymakers should “implement strategies to increase competition.” While such a pedantic recommendation may seem superficially appealing, there is only one small problem with New America’s analysis: Their own data do not support their argument.
As shown on the table below, New America’s 2013 Report includes pricing data from eight U.S. cities. Using the 2013 Report’s own data, I have summarized the names and the number of competitors in each U.S. city. Now, look at the final column of the table: In each city, there are “at least three competitive service providers” and, in some cases, more. Moreover, three of the eight cities include a municipal provider. If you haven’t caught on (and that’s no crime, since apparently the authors of the 2013 Report and others didn’t either), let me be clear: The data relied upon by New America’s 2013 Cost of Connectivity Report does not permit one to make claims about the U.S. being a laggard due to a lack of having “at least three competitive service providers,” since in all the markets they survey there are at least three providers. Even worse for their story, the correlation between the number of competitors and the lowest prices reported for each U.S. city for the triple-play package is positive (more competitors, higher prices). Once again, the analysis contained in New America’s 2013 Cost of Connectivity Report is inaccurate, careless and, thus, uncompelling.
Over my twenty-year career in telecommunications policy, I’ve seen a lot of international comparisons. What I rarely see is anything useful coming from them. Most of the time, the analysis is done poorly (see here, here, here, here, here, here, here) and when done well, it often is still unsatisfying. Comparing outcomes in different countries is tricky. To do it right, you must address demand-side conditions, supply-side conditions, and regulations. The 2013 Cost of Connectivity report does none of this detailed analysis, but merely compares price measures (poorly constructed) at face value.
An example reveals the problem with this approach. The 2013 Report compares Seoul, South Korea, to Chattanooga, Tennessee. Really? How can one plausibly claim that simply comparing prices between these two cities means anything at all? For one, population density in Chattanooga is about 900 persons per square mile; but population density is Seoul is 44,000 persons per square mile—a near 50-to-1 ratio. Accounting for household size, for every home a broadband network passes in Chattanooga (a key determinant of costs), it passes about forty in Seoul! Add that to the fact that the average income (GDP per capita) in South Korea is a little less than half that of the U.S. ($22,600 versus $51,700), and, despite that, the cost of living in Seoul is 50% higher than in Chattanooga. Now, to pile on, add profound cultural differences between the two cities. Advanced statistics performed competently by skilled analysts, not naïve price comparisons, are required for these types of comparisons, and even then it’s not always the case that the findings of the analysis are policy-relevant.
It’s not just me complaining. Congress explicitly rejects the type of naïve comparisons made in New America’s Cost of Connectivity Reports. In the Broadband Data Improvement Act, Congress outlined some relevant considerations when comparing broadband across countries, including, inter alia: population size, population density, topography, and demographic profile, market structures, the number of competitors, the number of facilities-based providers, the types of technologies deployed by such providers, the applications and services those technologies enable, the regulatory model under which broadband service capability is provided, the types of applications and services used, business and residential use of such services, and other media available to consumers. New America’s Cost of Connectivity Report considers none of these factors—to the contrary, they consider nothing.
I think everybody would love to pay $15 month for a triple-play service, but it’s not going to happen here. Programming costs for multichannel video are alone about $30 per subscriber per month, and more competition among broadband providers isn’t going to make that number go down. Comcast pays about $10 per video subscriber per month in interest expenses alone to fund its network. Also, municipal systems, which presumably are not profit maximizing, charge prices very much in line with the private sector firms they compete with, despite the fact they often operate under more favorable conditions (lower debt costs, better rights-of-way deals, cross subsidies) than do private firms. (In cases where it is unprofitable to privately provide broadband, such preferences may be sensibly exploited.)
There are some really hard and important policy changes needed for the U.S. telecommunications market. Our laws are outdated and our regulations are increasingly ridiculous runarounds of that outdated law. Conscientious research is desperately needed, and international comparisons, poorly done, are not very helpful.
New America Foundation’s poorly reasoned and executed Cost of Connectivity Reports are pure noise at a time when we need more signal.