May 8, 2014 2:48 PMletter to FCC Chairman Wheeler, complaining that the Chairman's proposal (described in his blog) to restrict bidding on at least 30MHz of the available spectrum in the upcoming incentive auction "is not how a market-based auction should function; it is how a cartel controls price." The House Republicans hit closer to the mark than you might.
Ironically, the purported reason for the restrictions is to prevent "one or two firms from running away with the auction." Such a result would be only be bad if it led to these "one or two firms" controlling enough spectrum to be able, at a later point in time, to exploit consumers through cartel behavior.
We know that cartels restrict output. If bidding restrictions, likewise, reduce output, then whose cartel tactics are likely to cost the consumer more?
The FCC's Theory on the Competitive Significance of Low-Band Spectrum
In his blog, the Chairman states that spectrum below 1Ghz is really important for commercial success in wireless. He believes this, presumably, because AT&T and Verizon (the two wireless companies with the most customers) also have more low-band spectrum than anyone else. However, correlation is not the same as causation.
Presently, here is how much total "low band" spectrum is available for commercial service:
Note that the chart above does not account for the broadcast spectrum to be auctioned in the upcoming incentive auctions. The FCC had originally speculated that the amount of 600MHz broadcast spectrum tendered for auction could be anywhere from 80MHz to 120MHz. The House Republicans speculated that only 60MHz would be tendered, due to the Chairman's decision to limit auction participation, and the value to broadcasters of surrendering spectrum.
If you want to see how the Chairman's plan will affect specific companies, the table below will give you an idea. This information is based on Table 18 from the FCC's 16th Wireless Competition Report (adjusted to reflect mergers), and it assumes that broadcasters will tender 84MHz to be auctioned. We also assume that the FCC wants to limit the amount of spectrum below 1GHz that any carrier can acquire; here, we use 1/3 of the post-auction total (73 MHz) as the limit.
Note, also, that in the above chart, neither AT&T nor Verizon's low-band spectrum comprises a majority of either company's total spectrum.
How Does the Chairman's Plan to Redistribute Low-Band Spectrum Effect Consumers?
The Chairman's plan is not just to limit the amount of low-band spectrum held by AT&T and Verizon. No, the plan also is designed to promote a more "equitable" distribution of low-band spectrum--at the lowest possible price to competitors of AT&T and Verizon.
These distortions are the primary reason no one expects the auction to recruit 120MHz of new low-band. The result of Chairman's bidding restrictions will be a 50% reduction in spectrum capacity available in this auction, and a total post-auction capacity restriction of almost 20% less total low-band spectrum available for U.S. consumers.
This last point is incredibly important. Restricting output is what monopolies do when they want to increase prices. Because consumer demand is fairly steady in the short term, the only way producers can move prices quickly is to restrict supply, which changes the equilibrium price to a point higher up the demand curve.
The Chairman of the FCC is unmistakably urging the Commission to adopt a plan that he knows will restrict output. The justification for this output restriction is ostensibly to prevent the top two firms from restricting output in some future time period.
What's the Worst That Could Happen?
If we assume the auction takes place with no bidding restrictions, reasonable spectrum screens, and we get active (but not maximum) broadcaster participation, then it seems possible that somewhere around 100MHz-110MHz in broadcast spectrum gets tendered. Moreover, let's assume AT&T and Verizon are allowed to buy as much as 60MHz-70MHz of the 100MHz.
Now, at some point in the future, the concern is that AT&T and Verizon will realize that demand is strong, every other competitor is capacity-constrained, and their opportunity to restrict output has finally arrived. If this day comes, and AT&T and Verizon decide, notwithstanding antitrust laws, that they want to maximize their opportunity, then they might look to the early 1970's OPEC.
As cartels go, early 1970's OPEC wrote the book on cartel coordination meeting exactly the right opportunity. As the world was already producing at maximum capacity, OPEC's 25% output reduction in November of 1973 changed the world.
So, for a worst case, let's assume that AT&T/Verizon will want to cut output by 25%. A 25% output restriction translates into somewhere between 40.5MHz and 43MHz, depending on whether you assume the two companies bought 60MHz or 70MHz of spectrum in this auction (25% of their combined new low-band total of 162MHz-172MHz).
What Does It Cost to Prevent?
This "worst case" outcome is, obviously, more than a little improbable. For the worst to happen, we have to assume: 1) AT&T/VZ would capture most of the profits from an output restriction, 2) both firms would/could disregard/circumvent the antitrust laws, 3) that such a steep restriction makes sense (25% is a lot), and 4) that the firms could effectively monitor and police their levels of capacity in service. Moreover, output "quotas" do not tend to work for very long (even OPEC members cheat on output quotas).
Nonetheless, the "worst case" does serve a purpose. In this case, it gives us some way of valuing the worst harm the Chairman's proposed bidding restrictions are supposed to protect us from.
If we know the economic costs of the worst case, we can assess the probabilities of that worst case, and get an idea of what preventing it is worth. So, here, the worst case is that consumers will face the higher prices that would result from an output restriction of about 40MHz of premium-grade, low-band spectrum.
But this is only a "risk"--it's not a certainty. But, even if you think there's as high as a 30% chance of the worst case happening, then we can assign a value on the worst case. In rough terms, it would be rational to engage in rules/regulations that "cost" up to 12MHz (in spectrum that will never reach the market) in order to prevent the worst outcome (i.e., a 30% chance of the economy losing the benefit of 40MHz of spectrum capacity).
Worth the Cost?
On the other hand, there seems to be a consensus among observers (both for and against the bidding restrictions) that the Chairman's proposed bidding restrictions will result in broadcasters bringing up to 40MHz less spectrum to the auction. But, even if the Chairman's restrictions "only" cause broadcasters to offer 20MHz less spectrum for auction, the loss is real and it is 100% certain.
Insurance is what the Chairman is selling with his proposed bidding restrictions. But, even at a Vegas blackjack table, insurance pays 2:1. At a guaranteed cost to the public of up to 40MHz, the Chairman owes taxpayers an explanation of why his bidding restrictions aren't the bad bet they look like.