Some of you may recall that the FCC had a big spectrum auction last summer. As is so often the case, the FCC and the usual cheerleaders for the incumbents could not wait to rush out and declare the FCC auction a success. I, contrarian that I am, observed that the cable companies and wireless incumbents had used the auction to kill DBS as a competitor (along with just about any other potentially disruptive incumbent). While no one could deny the DBS guys essentially dropped out by round 11 (and the Dolan Family, the other big potential new entrant, dropped out by round 20), no one could prove why. While I had a argued against letting cable operators even participate in the auction for fear they would block DBS and other operators, I couldn't prove it. Sure, they had motive and opportunity. Sure, the DBS providers had paid nearly $1 Billion “upfront” to play and clearly came committed. But these were, after all, hotly contested very valuable licenses. Was it just DBS amatures getting more than they bargained for at the bidding table? Or was it…..MURDER. O.K., not murder. But was it collusion, tacit or otherwise, among incumbents to keep out a disruptive player? Or was it just the auction working its magic?
Enter Dr. Gregory Rose. Last May, Dr. Rose (along with Mark Lloyd) did a study of ten years worth of FCC spectrum auctions for the Center for American Progress. His conclusion? Incumbents leverage the rules to keep license prices by signaling down and keep out new entrants by blocking.
Now Dr. Rose has done two follow up studies looking specifically at the AWS Auction. This invloved crunching over 16,000 individual bids to track how the bidders behaved. Quite the painstaking undertaking. To analyze signaling and retaliatory bidding, Rose used a methodlogy widely accepted in academia developed some years back by Peter Cramton and Jesse Schwartz. Cramton, it should be noted, helped T-Mobile lobby the FCC in the lead up to the AWS auction, so the incumbents should accept his methodology.
Rose's first report,
How Incumbents Blocked New Entrants In The AWS-1 Auction: Lessons For The Future, provides an in-depth analysis of the blocking strategies used by the incumbents and why they proved succesful. DBS Wireless, Dolan Family, Atlantic Wireless and Antares found themselves swarmed by incumbents every time they bid on a license. Amazingly, after the targeted new entrant dropped out of the bidding, most of the incumbents drifted away and the license bidding reverted to a standard contest among two or three bidders. Further, as Dr. Rose demonstrates, targeted new entrants faced specific opposition from a set group of opponents (Spectrum Co., T-Mobile, MetroPCS, and others listed in the report) at rates far higher than any other potential bidder. The anomolous behavior of incumbents and the unusually high (more than two standard variations from the mean) rates of collision between incumbents and the four targeted new entrants are most easily explained by a tacit or explicit agreement to block new entrants. any other conclusion, argues Dr. Rose, requires so many unlikely assumptions as to lie beyond the realm of probability.
Rose's second report, Tacit Collusion In The AWS Auction: The Signalling Problem, looks at the use of bids to communicate. In particular, Rose addresses a type of behavior called “retaliatory bidding.” As it sounds, retaliatory bidding works by convincing people not to bid against you because you punish them for it. Bidders only need to make a few retaliatory bids to establish their “auction cred” as mean sons-of-incumbents best left alone. In the Cramton-Schwartz Paper on the PCS Auction, for example, Cramton and Schwartz identified only 37 retaliatory bids — 23 of which involved the use of codes in a way prohibitted by the FCC under its current rules. But the 14 pure retaliatory bids out of nearly 19,000 Cramton and Schwartz identified produced a huge “indirect demand reduction” that allowed retaliating bidders to pay 40% less for licenses than non-retaliating bidders.
Using the Cramton-Schwartz methodology, Rose found 31 cases of retaliatory bidding in just over 16,000 bids, of which 13 were immediately succesfull (i.e., the target to withdrew from the contested license). But even unsuccesfull retaliation establishes a reputation as a bidder who will retaliate, and therefore one to avoid unless you have the cash and the will to defend yourself. Rose estimates that the rate of retaliatory bidding has increased since the Cramton-Scwartz paper, and that the results are more profitable for the retaliating bidders. Whereas retaliating bidders in the 1994 PCS Auction studied by CRamton & Schwartz paid 40% less for comparable licenses than non-retaliating bidders, Rose found that retaliating bidders in the 2006 AWS auction paid 41% less than non-retaliating bidders.
In both studies, Rose concludes that the FCC made a mistake when it failed to adopt the anonymous bidding rules proposed by then-FCC Chief Economist Leslie Marx. Under anonymous bidding rules, bidders see only the high bid for each license. They do not see what licenses others have bid on, or how much they bid. This denies them the information needed to target new entrants or signal each other.
To my great disgust, it was the Democrats that sabotaged anonymous bidding last time. (Never let it be said I don't stand up for Kevin Martin when he does the right thing.) Despite the fact that the
Department of Justice, the Federal Trade Commission, and a coalition of public interest groups all supported anonymous bidding, the Democrats sided with the “industry consensus” that anonymous bidding could derail the auction. So last April, I got to watch some Bizzaro World FCC in which Republicans Chairman Kevin Martin and Commissioner Tate warned against puting too much trust in industry consensus and protecting the market against collusion, but were caught in a two-two split against Democratic Commissioners Copps and
Adelstein insistance that FCC believe industry and not change the rules without solid evidence because of the possible unintended consequences of regulation.
As a result, the FCC adopted a compromise proposed by T-Mobile's Peter Cramton. Cramton argued that his study of the 1994 PCS auction failed to consider if the auction was sufficiently competitive and that, with more bidders per license, the AWS auction could be open without risking the collusive behavior of the 1994 PCS Auction. So the FCC adopted a rule that if there were, on average, three bidders per license they would have open bidding. But if the “modified eligibility ratio” were les than 3, than the FCC would use anonymous bidding. The Democrats grudgingly accepted this compromise.
Eyebrows should have shot through the ceiling last August when the initial eligibility ratio was 3.05. Wow! Just enough bidders to avoid anonymous bidding. But rather than suspicion, staff probably felt relief that they wouldn't have to worry about political flak from the Bizzaro Democrats.
Rose discusses this in his study, observing that the AWS Auction had 4 qualified bidders that never bid, and 7 bidders that bid once then never bid again (and did not win their licenses on the single bid). Take away these 11 bidders and the eligibility ratio would have dropped to 2.89 and the FCC would have used anonymous bidding. Did the incumbents put the 11 bidders up to it? Rose freely admits he has no proof, but that the math is certainly suggestive. Perhaps the Bizzaro Republicans that care about competition will investigate, if not blocked by the industry loving Bizzaro Democrats.
More seriously, as Gigi Sohn wrote in her blog, the FCC faces a defining moment in spectrum policy. It must decide what rules to adopt for the upcoming 700 MHz Auction. Even more than last August's AWS Auction, the 700 MHz Auction has the potential to completely redefine the wireless competition landscape. With rules that would promote competitive new entrants, the 700 MHz Auction could provide serious relief for a broadband market choked by a wireline duopoly and a wireless market crippled by a spectrum oligopoly. With the Rose studies, the Democrats should have the proof they need to adopt anonymous bidding, assuming the Republicans remain interested in doing what's right.