The Google Spectrum Gambit
Things have turned quite interesting in the upcoming Federal Communications Commission auction of new spectrum. Google announced today that it was willing to bid at least $4.6 billion in the upcoming 700 MHz auction, so long as the spectrum was sold subject to four open platform requirements. At first glance, that must seem odd: why does Google want restrictions imposed that it could simply implement as the auction winner? The point of course is that Google wants these requirements imposed on any winner, not just itself. Google might be happy to go into the spectrum business but it clearly believes that it benefits if open-platform spectrum is created. Google has effectively offered the FCC a price floor for the spectrum, so long as it gets the rules that it wants. What should the FCC do?
Adding the open-platform requirements will almost certainly reduce the revenues that the FCC will collect in the auction. Many of the incumbents may be discouraged from bidding under the open-platform requirements. But revenue maximization shouldn’t be the government’s chief concern: the government should be in the social welfare business and it should be willing to accept reduced revenue if it can buy enough telcom competition.
We need to take a step or two backwards. We are at the intersection of the fight over open vs. closed platforms and the role of the government in managing spectrum. If you have been reading this blog, you know that we see the open/closed fight frequently. The recent kerfuffle over the iPhone is the most recent example, where many are gnashing their teeth over the fact that the iPhone is available only with an AT&T wireless package. No iPhone for me as a Verizon customer (can you hear me now?). For some, open versus closed is almost a matter of religion, but there is no simple approach on this (see my recent book review of Invisible Engines and my published paper on copyright and technological contracts).
What does this mean for how we should auction the 700 MHz spectrum? We should start with auction revenues. If I own an unpainted barn, should I paint it red or blue before auctioning it? The answer is that I shouldn’t paint it, but should instead offer to paint it whatever color the winner likes. If I paint it red, some blue-lovers will drop, and the same is true if I paint it blue. Imposing the condition reduces the value of the auctioned item for someone who would prefer not to be subject to it.
In this case, imposing the four open-platform conditions sought by Google will discourage others from buying. That presumably is one of Google’s goals, as it frankly acknowledges in a July 9 filing that the unencumbered spectrum is probably more valuable to the incumbents. Part of what the incumbents undoubtedly invest in in buying additional spectrum is keeping out other competitors.
But revenue maximization shouldn’t be the government’s main goal in running the spectrum auctions. Switch to taxicabs and consider a toy model. Assume that we have an incumbent monopoly provider who will earn $100 profits in the next period. Assume consumer surplus of $10. Suppose that the government considers auctioning off a second taxicab license and that if we have a new entrant, we get an outcome of $40 in profits to the incumbent, $40 profits to the entrant and $40 in consumer surplus. What will happen in the auction for the license?
A potential entrant would be willing to pay up to $40 for the license. But the incumbent would be willing to pay more. If the incumbent paid $41 and put the license on the shelf, it would then earn $59 in profits in the next period. Far better to have the $100 to be sure, but if the government moves forward with another license, the incumbent would be better to buy it than to allow competition to emerge. The social welfare is higher if we actually create a second competitor, and yet the auction process won’t do that.
This is not just back-of-the-envelope analysis. Well it was, but Tom Hazlett and Roberto Munoz have looked at these issues more systematically in a paper doing a welfare analysis of spectrum allocation policies. They note the conflict between auction revenue enhancement and increasing competition and conclude that the government should often invest more heavily in competition even if that comes at the price of lower auction revenues. That doesn’t answer the open versus closed question, but it does tell me that the government needs to be critically aware of how auction design affects the overall level of competition.