Innovation Files has moved! For ITIF's quick takes, quips, and commentary on the latest in tech policy, go to

Flaws Remain in OTI’s “Cost of Connectivity”

The Open Technology Institute recently released the latest version of its “Cost of Connectivity” report. We at ITIF have repeatedly criticized past “Cost of Connectivity” reports for their flawed methodology (criticisms, by the way, shared with many others). The most recent OTI report continues this tradition, relying only on advertised broadband plans in a handful of cities. In keeping with this tradition, we offer the following constructive criticism in hopes that OTI will continue to improve their methodology going forward. One big improvement in this year’s report is the decision to drop the comparison of “Triple Play” bundles, with the recognition that the variation in cost and quality of programming bundles from country to country is too great to offer a meaningful comparison. Hopefully next year’s report will recognize the U.S. broadband market for the success it is and leave us with even less ammo for criticism.

It is not clear that we can draw any real conclusions from this year’s collection of data, given the tiny sample size and the disparity between advertised and actual speeds in Europe, not to mention the remarkable differences in history, culture, and industry structure. Although OTI has tempered some of their conclusions from this flawed set of data this go-around, statements like “U.S. municipal broadband providers offer some of the fastest speeds available in the country at relatively affordable prices” persist, despite being unsubstantiated even by their own data. If anything, OTI’s data shows that the US approach of lightly regulated facilities-based competition is doing a remarkable job, keeping our broadband speeds and prices competitive despite a number of disadvantageous this country faces in deploying broadband.

One problem with OTI’s report is its reliance on advertised speeds. Here we have good data based on SamKnows studies performed in both the U.S. and in Europe measuring actual broadband speeds. The same company used the same methodology on both sides of the pond, finding that American ISPs provide, on average, 101 percent of advertised speeds. Europe, on the other hand, has some work to do. On average their actual download speed is 75.6 percent of the advertised speeds, with xDSL speeds dropping as low as 44.3 percent of advertised speed in France and 45 percent in the UK (remember Paris and London “leading” in the OTI report?). Even with FTTx technology (which seems to be what OTI is fixated on) the EU only hits 84.5 percent of advertised speeds.

Beyond this basic problem of discrepancies between advertised and actual speed, reliance on advertised pricing data allows OTI to cherry-pick particular operators that are focused on a narrow, easy to serve geography. Larger U.S. operators cover wide areas with vastly diverse geographies, with generally uniform pricing. Reporting the advertised speed of a fiber carrier covering a low-cost, high-return area, such as parts of San Francisco for example, tells you very little about the appropriate federal broadband policy for the country as a whole.

To these broader policy concerns, even the flawed data set of the “Cost of Connectivity” clearly shows the value of dynamic, facilities-based competition. The U.S. cities that lead their ranking, Kansas City, KS and Kansas City, MO, are leaders because of Google Fiber working with those cities to lower the cost of deploying a fiber-based new entrant. Such a model, with local policy makers working with private operators to lower costs, points the way towards successful fiber deployment.

Contrast the success in Kansas City with the municipal networks OTI has so much affection for. It is clear that, even with OTI’s skewed data, the municipal broadband networks surveyed do no better than private networks. In fact, Bristol, VA, with its OptiNet municipal offering, comes in dead last in OTI’s ranking of the cost of a 25 Mbps connection. So much for the notion that eliminating profits from broadband companies results in cheaper broadband.

OTI’s data should be cause for celebration of our light-touch facilities based competition. But instead many appear to be drawing conclusions unsupported by the evidence. The myth that America is “falling behind” continues to be perpetuated, despite strong evidence to the contrary. I suppose such stories are easy to write (just repeat what Susan Crawford writes) and are probably popular reads, but policy-makers should rely on more rigorous analysis.

It is well established that the U.S. has some of the most affordable entry-level broadband. We do indeed pay more than some countries for higher speeds (which may explain the ire of the digital elite employed at some more liberal think tanks). But I would argue that progressive pricing is a fair way to spread the cost of deploying and operating expensive networks. The fact that our broadband industry has achieved competitive speeds while maintaining extremely low entry-level pricing is remarkable considering the hurdles we face with sprawling suburbs, rural states, low levels of computer ownership, and high rates of poverty. The OTI “Cost of Connectivity” report does nothing to justify the radical changes in policy it calls for.


Image Credit: Flickr User Kainet

Print Friendly, PDF & Email