BT Americas (BTA) has repeatedly beseeched the Federal Communications Commission (FCC) to revoke its regulatory flexibility granted to DS1 and DS3 special access – and now argues that all U.S. Ethernet services should be similarly regulated if they are to perform as well as those in Europe. Further, BTA often asserts that the performance of its Openreach subsidiary supplying special access services in the UK is far superior to U.S. performance. While BTA’s past claims have been repeatedly refuted by AT&T, its recent Reply Comments introduce a new “study” purchased from European telecom consultancy, WIK. This study argues that Ethernet prices are distinctly higher in the U.S. than in Europe, and that this unhappy state is due a lack of competition and competent regulation in the U.S. A little detective work shows these claims about Ethernet to be as fanciful an overreach as BTA’s prior claims about DS1 and DS3 services.
WIK’s February 2016 report presents two sets of Ethernet price “data.” The first comes from another consultancy, Ovum, and dates from August 2013. WIK uses these Ovum (2013) data to estimate Ethernet prices for the U.S., the UK, France, Germany and the Netherlands. Based on this comparison, BTA and WIK conclude that Ethernet prices are higher in the U.S., and that Ethernet adoption is lagging in the U.S.
When drawing this conclusion, it is unfortunate that BTA and WIK apparently were unaware that on Sept. 28, 2015 Ovum published an update to its 2013 report. Examination of Ovum’s 2015 report reveals major changes in the Ethernet market over the intervening two years. In 2013, Ovum reported U.S. prices for 1G metro Ethernet to be $31,060 per year. Just two years later, this price had dropped by 36% and was $19,879. Declines in U.S. 10G service prices were even greater – dropping by nearly 58% from $61,237 in 2013 to $25,873 in 2015. But even more telling, Ovum’s 2015 report projects that by the 2018-20 period, U.S. prices for 1G and 10G Ethernet service will be no higher (and sometimes significantly cheaper) than the prices from the four European champions that WIK selects as comparators to the entire U.S. Further, over the 2015-20 time period, Ovum (2015) projects that U.S. metro Ethernet lines will grow at a compounded annual rate of 16.9%, while the four-country European growth rate will be only 9.2%. The moral here, if you rely on stale data, expect rotten results.
Perhaps aware of the frailty of using old Ovum data, WIK and BTA propose an alternative data set of “rack rates” developed in-house by WIK. These Ethernet circuit rates are developed from tariffs filed by European telecom operators and from suggested prices printed in “service guides” published by AT&T and CenturyLink for the U.S. Again, BTA and WIK pronounce Ethernet to be more costly in the U.S. than in Europe. But such “rack rate” comparisons are meaningless unless these rates have the same relation to as-paid rates in all of the compared countries. But as even BTA and WIK admit, they do not. WIK notes that in Europe, regulations generally require Ethernet to be offered under tariffs that do not permit discounts. Thus, outside of some limited pockets of deregulation, the European tariffed rack rates are what customers must pay. This is not the case in the U.S., where the relationship between service guide-suggested prices and as-paid prices is akin to the relation between nightly rates listed on the back of your hotel room door and the rate you actually pay.
Finally, WIK returns to Ovum’s 2013 data to create its Figure 19 (reproduced below), that it claims sums up the relation between U.S. and European Ethernet pricing.
Figure 19: Effective unit price per 10Gbit/s metro Ethernet end points in France, Germany, the Netherlands, the UK, and the United States (unadjusted 2013 USD).
But if WIK had used Ovum’s 2015 data, its Figure 19 would have looked quite different – and not at all suggestive of U.S. Ethernet prices that are too high or in need of any European-style regulatory correction to ensure their competitive nature.