Restoring competitive balance to the upcoming spectrum auction
July 23rd, 2007 | Published in Google Public Policy
In recent days, and especially following Eric Schmidt's July 20 letter to FCC Chairman Kevin Martin, many people have asked us a straightforward question: why don't you just attempt to win the spectrum bidding outright, and then implement an open wholesaling business model yourself? Or, as AT&T has put it, "put up or shut up."
That question makes a lot of sense, especially for those most familiar with the ordinary marketplace structures they see on eBay, or Home Shopping Network, or at Target. In those everyday cases, the buyers and sellers collectively determine what is the fair market value for something, based on what willing participants on both sides agree should be the price. The free market is the optimal market.
But an FCC spectrum auction is a very different animal. Unlike in most commercial transactions, participants in an FCC auction operate in an artificially defined market. Initially, the spectrum comes sliced and diced in predetermined packages of varying bandwidth, geography, and duration. Those discrete slices tend to be compatible with what regulators perceive to be the prevailing services and technologies of the moment, such as centralized voice communications.
Further, given the sizable investments involved, only well-capitalized corporations can afford to bid at auctions. Ordinary citizens or entrepreneurs with novel ideas don't even show up. Even so, players still come to the table with unique assets, and in some cases disparate business models. For the upcoming 700 MHz auction in particular, the issue boils down to the different incentives at work between the existing national wireless carriers -- the incumbents -- and those companies seeking to enter the market for the first time -- potential new entrants.
As we have seriously considered entering the 700 MHz auction, we have been consulting with auction experts and game theorists to help us better understand the dynamics of a typical spectrum auction. What they have been telling us is that in a head-to-head bidding war between an incumbent wireless carrier and a potential new entrant, the incumbent almost invariably will prevail. Why? The answer involves two key economic factors: what we call the "incumbent blocking premium" and the "incumbent dilution discount."
Incumbent blocking premium
A significant economic factor that comes into play in an auction environment is the incentive and ability of one of the players to thwart the designs of the other. In the context of an FCC spectrum auction, there are at least three pertinent elements.
First, incumbent wireless carriers come to the auction with a vast array of existing assets, including thousands of radio towers, tens of thousands of miles of communications "backhaul" networks, and millions of customers, along with numerous retail outlets and tons of advertising. And perhaps most important of all, incumbents already own lots of spectrum -- much of which the FCC gave away for free some years ago, rather than sold at auction. By contrast, a true new entrant has none of these assets. Thus, to an incumbent, purchasing another wireless license is just an incremental investment, one made that much less costly given the existing, readily-available business inputs. To a new entrant, facing the daunting challenges of actually building and operating a network for the first time, the investment is less certain.
Second, the incumbent has the added benefit of operating in a less than fully competitive environment. Some use the term "monopoly rents" to describe the situation where a company enjoys revenues and profits that exceed what normally would be the case in a robustly competitive environment. Potential new entrants do not enjoy the same advantage.
Third, and perhaps most important, the incumbents have every incentive to preserve and protect their existing business model. Given their investment in all the necessary business inputs, and the relatively high prices and low bandwidth characteristics of their existing service offerings, the incumbents must prevent the entry of potential competitors to the market. In a spectrum auction, this means paying whatever it takes to block new entry. Not surprisingly, economists call this a "blocking premium."
As the diagram below shows, these three elements together mean that an incumbent will almost always be in a position to outbid a potential new entrant, simply by bidding above and beyond the fair market price. Unlike in a healthy auction situation, the final price would be higher than otherwise would be commercially reasonable.
The second significant economic factor that comes into play in any auction environment is related to the number of players who actually show up to participate in the bidding. Again, in a normal commercial environment, market prices are shaped by the number of willing buyers, from just a few to potentially millions. However, an FCC spectrum auction presents a comparatively artificial scenario.
Given the existence of the incumbent blocking premium, as described above, it is often the case that there are no new potential entrants to bid against an incumbent. Why should a new player even bother to bid, or bid aggressively, if the incumbent inevitably will prevail? Where there may be only two incumbents -- or even one -- left to bid for a license, obviously the resulting price will not reflect the fair market value that otherwise would have been reached. The dilution of competitive bidders means the final price will be lower than otherwise would be the case. Recent studies have confirmed that this is a pervasive aspect of the FCC auction environment. 
Un-skewing the spectrum auction
When looking at the combined impact of the incumbent blocking premium and the incumbent dilution discount, it's easy to see how FCC auction results become skewed.
Ironically enough, it is Google that has been accused of attempting to skew the auction structure, by our recommendation that the licenses be conditioned on certain "open platforms" requirements. Of course, as we have explained it is the current auction system that skews the results away from potential new entrants and in favor of existing incumbents.
Our position is simple enough. FCC Chairman Kevin Martin and the other commissioners have argued persuasively that we need a real third pipe broadband competitor in this country. They also believe that the upcoming 700 MHz auction is the best way to get there. All we are saying is that, based on what we know, new broadband competition will emerge from the upcoming auction only if the FCC's rules allow it to happen. For Google, and other potential new entrants, the prevailing imbalance can be corrected most effectively by introducing license conditions based on open platforms.
While Google embraces the kinds of openness and innovation that are the hallmark of the Internet, the incumbents apparently prefer their existing business models. That of course is their prerogative. However, open platforms -- specifically, open applications, open devices, open wholesale services, and open network access -- together make the spectrum more valuable to Google, or any other potential bidder seeking to create innovative, higher-speed, lower-priced offerings.
That is why Google has indicated that it is willing to spend a minimum of $4.6 billion in the auction, which is the FCC's reserve price for the particular spectrum block in question. At the same time, incumbents are unable to leverage anti-competitive blocking in this scenario. Regardless of who wins the bidding, however, the end result is an auction that yields a fair market price, with the added bonus of a new broadband network that is open to all comers. The American people get full value for their spectrum, plus open broadband platforms -- and even the possibility of a real third pipe competitor. Not a bad deal overall.
