Has the Federal Communications Commission finally learned its lesson on spectrum management?
The FCC began proceedings yesterday that could OK Dish Network's plan to use existing spectrum to build a terrestrial 4G LTE mobile broadband network. The rulemaking follows the agency's earlier rejection of Dish's request for a waiver of license conditions, which prohibit using the spectrum for anything other than satellite-based applications.
The decision to proceed with the slower but more formal process was certainly motivated in part by the recent fiasco involving LightSquared. In January 2011, the FCC granted LightSquared a waiver similar to the one requested by Dish, also to build a ground-based mobile network using spectrum currently limited to satellite applications.
That decision has been cursed. After the waiver was granted, both government and private parties complained to the FCC that LightSquared's spectrum sits too close to bands used for some GPS devices and would likely lead to interference. In February of this year, the FCC reversed itself, suspending LightSquared's waiver indefinitely. The company, which has already spent $4 billion building its network, has since lost its CEO and key partners, and faces an uncertain future.
Meanwhile, congressional Republicans raised concerns that the initially favorable treatment of LightSquared was motivated by undue and potentially illegal political pressure from the White House. So far, the agency has refused to release documents related to the waiver demanded by Sen. Chuck Grassley (R-Iowa), who has placed holds on two pending nominees to the FCC. The FCC is now operating with a bare minimum of only three commissioners, and the stalemate has no end in sight.
Both the LightSquared and Dish proceedings highlight the growing chasm between the needs of mobile consumers and the FCC's plodding and often politicized processes for spectrum management. If the mobile revolution is to continue without interruption, Congress will have to make radical changes to how the agency operates.
Micromanaging spectrum doesn't work
Yesterday's rulemaking reflects the FCC's admirable efforts to satisfy exploding demand for mobile broadband. Since the introduction of the iPhone in 2007, mobile use has grown at an unprecedented pace. According to the FCC's 2010 National Broadband Plan, the agency must allocate an additional 300MHz of usable spectrum by 2015 to head off serious network congestion and the possibility of slowdown for the fast-growing mobile services industry. Reallocating satellite spectrum for a new 4G LTE network would be a significant step toward avoiding an imminent "spectrum crunch."
But while a formal rulemaking minimizes the kinds of procedural and political issues plaguing the LightSquared deal, it also means Dish will have to wait several months or longer before beginning build-out of its network. The rulemaking may not be concluded until the end of this year or later, depending on responses the agency receives for its requests for public comment on a wide range of technical issues. The delay serves neither consumers' appetite for more bandwidth or the FCC's goal of expanding mobile competition.
The real problem is the FCC's licensing system. Under decades-old policy, the agency assigns spectrum to different applications and geographies through licenses that severely restrict how and by whom different frequencies can be used. (Today, the agency manages over 50,000 licenses, and doesn't even have a complete inventory.) The restrictions, however, are designed more to shape new industries to the FCC's idiosyncratic vision of what's best for the "public interest" rather than to minimize technical problems such as interference.
This command-and-control strategy can't keep up with technology and consumer needs that change faster all the time. As a result, the FCC has become a bottleneck in effective network design and management. As FCC Commissioner Robert McDowell wrote in Wednesday's proceeding, "The Commission has a checkered past of micromanaging spectrum use only to find years later that technical innovation and market demands have evolved past the government's myopic view."
Regardless of the outcome of the proceeding, collisions between innovation and outdated FCC policy are certain to accelerate. New uses and devices for mobile services continue to develop at an accelerating pace. But the FCC's capacity to adjust existing licenses and allocations is limited and slow. The LightSquared fiasco and the delay of Dish's efforts to build a competitive mobile broadband network with existing spectrum underscore McDowell's concerns.
Assigning property rights to spectrum
We need a long-term solution that facilitates rather than impedes rapidly changing demand for our valuable and limited spectrum resources. Fortunately, a better approach exists. And oddly enough, it was first proposed over 50 years ago, in a seminal article by economist Ronald Coase. The article, titled simply "The Federal Communications Commission," was published in 1959. Its findings were key to Coase's being awarded the Nobel Prize in 1991. (Coase recently celebrated his 101st birthday.)
Coase argued that instead of granting inflexible licenses based on squishy "public interest" priorities--often warped by political influence--the government should simply auction spectrum to the highest bidder. Under Coase's plan, the auction winner would be granted full properties rights to specific frequencies. The owner could then use the spectrum however they saw fit, and could later resell it without government oversight or approval. The government would only step in if serious problems of antitrust arose.
(To ensure government agencies did not hoard spectrum better used by consumers, Coase argued that federal agencies should also be required to pay for their spectrum allocations, even though payment would effectively be to another part of the government. That proposal has never been put into effect; today, federal agencies hold vast swaths of unused or underutilized spectrum, bearing out Coase's concerns.)
Spectrum auctions and property rights represented a radical alternative to the system that had been in place even before the FCC was founded in 1934. Before and since, licensees paid little or nothing for the exclusive rights to frequencies. Allocations are based on transient and often unarticulated views by the agency of what is best for the public interest.
Licenses, however, come with severe limits. Once allocated, frequency allocations are locked into specific applications that often became obsolete. Since the FCC rarely refuses to renew licenses, spectrum became splintered and increasingly inefficient. Without property rights to spectrum, secondary markets for transferring licenses play a limited role in speeding up the change from old to new technologies. For one thing, the FCC must approve all transfers. Even when they do, the new licensee is still bound by all the old restrictions.
Under Coase's plan, spectrum holders wouldn't need a waiver or a formal rulemaking to resell their allocation or use it for new applications. Consumer demand would determine the best use of limited resources, just as it does in unregulated industries. If Dish or LightSquared had an unrestricted right to use their allocation of spectrum to best serve their customers, for example, none of the costly delays and expensive machinations currently gumming up the works would be necessary.
Despite these limits, Congress and the FCC ignored Coase's proposals until 1994, when the FCC finally began the process of auctioning spectrum instead of simply giving it away. But the auction winner still only gets a limited-use license. The agency also has a hard time shaking old habits. It often limits auction eligibility to shape the competitive landscape of emerging industries, and attaches unrelated conditions to licenses. Both practices limit the ability of winning bidders to transfer the license to a future user who might put it to better use.
A market-based solution to possible interference
One argument for the FCC's command-and-control system is that it ensures the agency has the ability to police potential interference issues, protecting investments by network operators and consumer device manufacturers. But Coase's property rights proposal also took into account the potential problem of conflicting uses and potential interference and proposed an efficient and elegant alternative.
Reviewing the early history of radio, Coase acknowledged that without property rights or regulation of any kind, it wasn't long before chaos reigned. But he also demonstrated that licensing tiny slivers of spectrum and severely limiting its use or transfer was an expensive and inefficient solution to potential interference.
Instead, Coase argued that once property rights were established and initial allocations set by auction, interference issues would be resolved just as other problems of conflicting property rights had always been resolved. The parties would start by trying to negotiate a solution that minimized damage being caused by either or both parties to the property of the other. If negotiations failed, the legal system waited in the wings, with its established rules of liability developed through precedent. The backstop of the courts would encourage the parties to settle.
As Coase argued here and in later articles, enforceable property rights ensure that the party that most values its current activity will win in any dispute over interference. In the classic example, Coase showed how the law had developed default rules to determine when railroads were liable to neighboring farmers for crop damage caused by sparks. If the railroad was liable, it would either pay for the damage or take steps to avoid it, whichever was cheaper.
If the railroad was given the right to pollute, Coase argued, the result would be the same, except that the farmer would have to pay to stop it--but only if that was cheaper than the cost of the damage. Either way, property rights and a clear rule of liability ensure that the net cost to the economy is minimized, and without incurring the added cost of an expensive and potentially corruptible regulator.
Indeed, the LightSquared problem is a perfect example of Coase's theory. LightSquared argues that its network would only interfere with GPS devices that haven't been carefully designed to ignore signals outside their allocated range. Regardless, whether LightSquared needs to restrict its network to accommodate those devices or whether GPS manufacturers need to redesign future devices to be more selective would be resolved through negotiations between the parties. Coase argues that with clear property rights, the most efficient solution would ultimately prevail.
Under a property rights model, in other words, it isn't necessary for the FCC to determine if LightSquared or the GPS manufacturers is causing the problem. (So far, the agency has sided first with LightSquared and now with the GPS makers, leaving everyone worse off.) If the parties failed to reach a negotiated solution, standard liability rules would allocate the costs based on relative fault, approximating what the negotiations had failed to establish.
The property approach is neither perfect nor costless, but Coase demonstrated its superiority over the FCC's longstanding system of regulation based largely on the undefined "public interest."
Instead, Coase believed that interference rights should be delimited in part by regulation and in part by standard liability rules. The role played by each, Coase argued, "can be answered only on the basis of practical experience." But, he continued, "There is good reason to believe that the present system, which relies exclusively on regulation and in which private property and the pricing system play no part, is not the best solution."
The FCC's expensive bungling of LightSquared represents the "present system" at its worst. Even at its best, a system that limits uses, excludes bidders, and resolves conflicts entirely through regulation and adjudication by a single federal agency has proven an inefficient and anachronistic way to manage spectrum. The last 50 years have only underscored the wisdom of Coase's proposal to take the FCC out of the day-to-day oversight of ever-expanding uses for spectrum and potential conflicts between users.
Treating spectrum more as property and less as the whim of a sluggish FCC is key to ensuring future innovations have the chance to take their rightful place in the mix of wireless services. Slowly but surely, the FCC and Congress have accepted many of Ronald Coase's recommendations. If new mobile technologies have any hope of succeeding in the next 10 years--let alone the next 50--we'll need to speed up the process of adopting the rest of them. And soon.