Many of you have probably heard about Google Fiber, where the search giant provides 1,000Mbps Internet service to homes and businesses in some areas of the country. Because average Internet speed in the U.S. at the end of 2014 was ~11Mbps, that makes Google’s $70/mo offering around 100 times faster. While only available in a few cities, Google Fiber is interesting both because of its technological superiority and because consumer satisfaction with these incumbent Internet providers is very low. (Comcast recently won Consumerist’s 2014 Worst Company of the Year award.) There are also some smaller and less well-known competitors in local areas, like our own LightSpeed here in Lansing, MI, that offer similar service.
What’s interesting here is the response from incumbent carriers when a competitive service is announced. When Google announced it planned to offer its service in Atlanta, Comcast responded by announcing it would also offer gigabit speeds in that city. Google Fiber was announced for Kansas City and Austin, and AT&T matched that offer. Most recently, Google Fiber announced plans to offer service in Charlotte, and Time Warner Cable announced huge no-cost upgrades in their service there.
If you think this feels a bit like predatory pricing by an incumbent monopolist, you’re not alone. But is what these companies are doing actually illegal under anti-trust laws? Unsurprisingly, this is somewhat complicated. Since the text of the law itself isn’t very specific, the details have largely been left up to the courts, and in the past few decades the libertarian-leaning “Chicago School” has been very influential in the federal judiciary. In general, this means it’s rather difficult to win an anti-trust case. It would be even more difficult if federal judges (who, with lifetime appointments, are generally older) apply industrial age economics to an information age problem.
Generally, in order for someone like Google to win an anti-trust case against these incumbent ISPs, they’d have to show that both (1) the ISPs reasonably calculate that their response will deter other firms from entering the market, therefore increasing their own profits, and (2) that the prices they’re offering in response are below “cost”–usually a specific type of cost called average variable cost, or AVC.  I think there’s a pretty good argument for #1, but #2 is a harder sell because of the way costs are structured in telecommunications.
But this is a damned-if-you-do, damned-if-you-don’t sort of situation, because it means that one of the following must be true. Incumbent wireline broadband ISPs either-
- Can increase bandwidth by 100x without raising prices and still be profitable. (But they aren’t doing it, e.g., because of insufficient competitive pressure.) or
- Are probably engaging in illegal predatory pricing because they are offering service below cost to deter potential competitors from entering the market.
This is a problem because barriers to entry in this market are high. Because the FCC under the Bush (W) administration got rid of rules requiring infrastructure sharing, entering this market means running your own wires all over town. Obviously this is very expensive, and requires extensive coordination with cities and other utilities to get access to rights of way and, e.g., utility pole attachments. This might be addressed by having the municipality run telecommunications lines with the rest of their other infrastructure, but in many cases state laws prevent municipalities from offering an alternative themselves. We in Michigan have our own version of this law at MCL §484.2252.
 Phillip Areeda, Louis Kaplow & Aaron S. Edlin, Antitrust analysis: problems, text, cases (2004); Herbert J. Hovenkamp, Predatory Pricing under the Areeda-Turner Test (2015), http://papers.ssrn.com/abstract=2422120 (last visited Apr 13, 2015).