Friday, May 10, 2013

Maximizing the Benefits of Future Spectrum Auctions

            Sponsored researchers already have entered the conversation about spectrum policy with a new objective of thwarting any effort to promote access by non-incumbents, or at least any carrier other than AT&T and Verizon.  These researchers will prove that denying incumbents the opportunity to acquire even more spectrum will reduce the government’s take.  I agree and empirical evidence supports this.  When the FCC imposed requirements of open access or sharing with first responders on a spectrum block, the amount bid was lower than unencumbered spectrum.

            But of course sponsored researchers want to extrapolate from this truth to many conjectures including the premise that any spectrum set aside would prevent the most efficient providers from doing more with more.  Somehow if AT&T and Verizon do not capture the lion’s share of any and all available spectrum, then both taxpayers and wireless consumers suffer.

            This premise does not pass a basic smell test.  We should appreciate that what the government takes now in spectrum auction proceeds, it loses in future tax revenues, because carriers can use their spectrum investments as offsets against income.  

            Perhaps more importantly we should consider what the two incumbents with over 70% market share can do with additional spectrum.  In the best case scenario they will put the spectrum to immediate use and abate any real scarcity.  In the worst case access to more spectrum eliminates incentives to more efficient use including the possibility of buying simply to deprive competitors of access and to preempt market entry.  Additionally incumbents possibly can “warehouse” the spectrum by not using it, but preventing other carriers from putting it to efficient and immediate use.

            Consider a commercial aviation analogy.  Let’s assume a highly congested airport can offer additional landing and takeoff slots, a result when an additional runway gets constructed, or when regulators relax a cap, or allow late night operations.  In this particular aviation market one carrier has a dominant market share, something that regularly occurs when that market represents a carrier’s hub, e.g.,  Washington Dulles for United; Philadelphia for U.S. Airways, Detroit for Delta and Dallas Fort Worth for American.  Dominant carriers have market power in their hub markets as evidenced by their ability to charge higher fares than cities with competitive commercial aviation markets. 

            These carriers will do anything to maintain their dominance including acquiring as many new landing and takeoff slots as possible.  Of course they will frame their acquisitions as serving the public interest and consumers, even if they have to use smaller aircraft—with less seat capacity—to ensure that every slot gets used.  With more available slots incumbent air carriers might determine that they will oversupply seating capacity with large planes.  But rather than pass on the opportunity to control even more access to the market, these carriers will acquire new slots at any price simply to prevent existing or prospective competition from flourishing. 

            In the short run everything looks grand: the government accrues higher auction revenues than contemplated, because of the market preemption benefits reflected in a dominant carrier’s win at all costs bids.  But in the immediate term consumers suffer from higher rates available to the fortress hub carrier.  Recently even corporate flyers have complained about the consequences of hub dominance and the reduction of competition and flight options in non-hubs.  In the longer term the tax benefits to incumbents and the elimination of most competitive benefits weigh in.

            Bottom line: if the FCC seeks to maximize short term spectrum auction proceeds it will guarantee that incumbents acquire most newly available spectrum further concentrating the market and reducing the benefits of facilities-based competition.

No comments: