Ownership concentration and strategic supply reduction

Main article

Two-sided auctions can be a useful tool for government agencies looking to reallocate scarce resources, such as electromagnetic spectrum, to more valuable uses.

In March 2016, the Federal Communications Commission launched one such auction, the most complex spectrum auction it had ever conducted. 

The agency was seeking to solve a bedeviling problem. The rapid growth in data and video usage by smartphone users had significantly increased the demand for mobile broadband spectrum. Much of that spectrum, however, was already allocated: in 2012, more than 8,400 TV stations held licenses for a six MHz block of spectrum that allowed them to broadcast programming to local customers with antennas. But fewer and fewer Americans were watching broadcast television via antenna. To best serve the country’s needs, therefore, the FCC needed to find a way to reallocate spectrum from TV stations to wireless carriers. 

The FCC wanted to use an auction to facilitate the reallocation—and for good reason. In comparison to other approaches, such as bilateral negotiations or take-it-or-leave-it offers, auctions rely on voluntary participation, encourage competition, and are relatively robust to legal challenges. 

In March 2016, the Federal Communications Commission launched one of the most complex auctions ever conducted.

After a long design process, the FCC came up with a clever two-sided auction: TV stations would bid to relinquish their licenses to the agency in a reverse auction. The FCC would then “repack” the remaining stations into different portions of the wireless spectrum, freeing up a large, contiguous block of spectrum that it would, in turn, sell to wireless carriers in a forward auction. 

Ultimately, the auction worked largely as intended. When it closed on March 30, 2017, the forward auction raised $19.6 billion in proceeds from wireless carriers, covering payouts of $10.1 billion to TV stations in the reverse auction, and raising $7 billion in proceeds for the U.S. Treasury along the way. Through the process, the FCC was able to repurpose 84 MHz of spectrum from broadcast TV to mobile broadband usage—less than the 126 MHz it had originally hoped, but still a substantial step towards alleviating the looming “spectrum crunch.” 

We regard this auction as a triumph of modern market design. It reallocated spectrum in a socially valuable way and simultaneously raised significant revenue for the government—all through a nonmandatory, market-based process. However, in a paper in the American Economic Review, we identify the joint ownership of TV stations as an important factor in the outcome of this auction. We also make suggestions for how to account for the joint ownership of TV stations in designing future auctions.

In our paper, we examine multi-license owners—firms that hold licenses for multiple TV stations. We find that owners of multiple TV stations have an incentive to strategically withhold some of their stations from the auction, thereby driving up prices for their remaining stations. Using a large-scale simulation exercise, we show that this strategic supply reduction can increase payouts to TV stations by as much as 42.4 percent. What is more, the approach of withholding licenses may have resulted in less available spectrum for the FCC to sell to wireless carriers—a loss for American consumers.

We see two ways the design of the auction could be adjusted to take into consideration the joint ownership of TV stations more fully. One suggestion involves a slight change in auction rules regarding multi-license owners. The other involves relaxing a particular constraint in the repacking process. Each one, we show, would help reduce the incentive for multi-license owners to engage in strategic supply reduction. 

Multi-license owners had an advantage in the FCC Incentive Auction

As noted above, the FCC’s auction had two stages: the reverse auction, in which TV stations bid to relinquish their licenses, and the forward auction, in which wireless carriers bid for portions of the newly available spectrum. TV stations that ultimately relinquished their licenses could opt to either go off the air altogether, move to a lower-frequency part of the spectrum, or share a channel with another station. 

Ultimately, the auction worked largely as intended, raising $19.6 billion in proceeds from wireless carriers.

Between these two stages sat the all-important repacking process. For the FCC to have a contiguous block of spectrum to sell to the wireless carriers in the forward auction, it had to find channels for all the stations that opted to remain on the air.

Repacking is extremely tricky, especially in highly populated areas with many TV stations. Depending on local geography, as well as the height and power output of their broadcast towers, two nearby stations can interfere with one another if they are broadcasting on the same channel or on immediately adjacent channels. So not only did the FCC have to clear an uninterrupted block of spectrum, but it also had to consider interference: the agency could not reassign, say, one station in Philadelphia and another in Harrisburg to channel 51 (or even, potentially, to channels 50 and 51). 

The upshot of these two concerns—interference and the need for a contiguous block of spectrum—was that some licenses were more valuable to the FCC than others. Much like a single holdout homeowner can derail an entire highway project, certain license holders had particular importance in the auction, because they interfered with many other stations around them. The FCC wanted these stations either to go off the air or to take one of other relinquishment options, in order for the rest of the repacking process to work. What the FCC was willing to pay for a particular license therefore depended on the value of the TV station as a business and on its interference constraints. The FCC formalized these considerations by assigning each TV station a broadcast volume, which scales the population served by the TV station by the number of its interference constraints.

The design of the reverse auction rested on an important assumption: that a license holder will make a “truthful” bid in the auction based on the real value of the TV station as a business. For a single-license holder, bidding truthfully is always the dominant strategy.

Owners of multiple TV stations have an incentive to strategically withhold some stations, thereby driving up prices for their remaining stations.

However, if a company owns multiple licenses, this assumption falls apart. Multi-license owners will naturally seek to maximize their total payout across all licenses, rather than the individual payout on each license. Hence, multi-license owners have an incentive to manipulate the auction. By strategically overbidding on some of their licenses or withholding them from the auction entirely, multi-license owners can, essentially, create scenarios where their remaining licenses become a bit like that lone holdout homeowner derailing the highway project. As a result, the FCC will be forced to pay more than the licenses would otherwise be worth.

Multi-license ownership is common in the United States Broadcasting Market

In our paper, we show that multi-license ownership and strategic supply withholding are not merely theoretical concerns. We examined all the auction-eligible stations in the continental U.S. and found significant ownership concentration. In 2015, the 1,670 TV stations in our sample that broadcast in the UHF part of spectrum were held by just 482 owners. Of these owners, 302 held one TV station across the U.S., 66 held two, 33 held three, and the remaining 81 owners held at least four stations. 

We also examined ownership patterns within designated market areas (DMAs), the local media markets to which the FCC assigns each station. Most DMAs had at least one multi-license owner, meaning that at least two stations within the area were jointly owned. Of note, multi-license ownership was prevalent in the highly populous DMAs that were especially important in the reverse auction.

Ownership concentration has traditionally been a concern for regulators, including the FCC. However, the agency’s regulations were mostly designed to prevent owners of major TV stations from unfairly influencing advertising rates. In other words, the FCC was trying to prevent a single owner from gobbling up (for example) the local ABC, CBS, and NBC affiliates in a single DMA. The agency’s rules did not apply to non-commercial, low-power, and satellite stations, or those that were financially distressed—some of which, despite their relatively low values as businesses, ended up being important in the reverse auction.

Indeed, as plans for the auction began to take shape, private equity firms began buying up local TV stations. From 2011 to 2015, three private equity firms—LocusPoint Networks, NRJ TV, and OTA Broadcasting—acquired 48 stations for at least $380 million. Most were failing or seriously distressed, and, crucially, most sat in major DMAs in the Northeast and along the West Coast. 

We show that strategic supply reduction by multi-license owners can substantially increase payouts.

Initially, onlookers speculated that these private equity firms were seeking to flip the stations in the auction. But our analysis suggests the private equity players had more sophisticated plans. We discovered that they relinquished only 19 of the stations they had acquired to the FCC in the auction and sold another 26 stations soon after it ended. In general, while these three firms made substantial profits on the TV stations they ceded in the reverse auction, they incurred losses on those they sold soon after. 

To us, this pattern seems far more consistent with strategic supply reduction than with simple flipping. In fact, we estimate that one of the firms made a 509 percent return on investment overall—a degree of profit that, we believe, would simply not be possible unless firms were engaging in some form of highly strategic bidding. 

The high cost of strategic supply reduction

While the real-world data gives us suggestive evidence of how strategic supply reduction played out in the FCC auction, we wanted to measure the likely effects of the strategy more precisely. 

To arrive at that more precise measurement, we created a simulation of the auction. First, we estimated each TV station’s value as a business, considering factors such as advertising revenue, transmission fees, and fixed costs. Then, we re-ran the auction, using the same rules as the FCC. 

We measured how the auction would have played out under truthful bidding and under strategic bidding. In the case of truthful bidding, we assumed each license was held by a single owner seeking to maximize their payout on that individual license and bidding truthfully based on the value of the business as we had estimated it. In the case of strategic bidding, we imagined a landscape closer to real life, in which some players in the auction were multi-license owners seeking to maximize their total payout across all licenses. We focused on equilibria of the ensuing game between the multi-license owners. 

Our simulation shows how significant the effects of strategic supply reduction can be. We modeled several different scenarios that considered factors such as the clearing target, which is how much spectrum the FCC sought to recapture in the auction. 

The approach of withholding licenses may have resulted in less available spectrum for the FCC to sell—a loss for American consumers.

Under one scenario, in which the FCC wanted to repurpose 126 MHz of spectrum, strategic bidding by multi-license owners increased nationwide payouts in the reverse auction by 42.4 percent. Assuming a lower clearing target of 84 MHz—what the FCC achieved in the real auction—strategic bidding increased nationwide payouts by a still substantial 13.5 percent. 

Interestingly, although these effects are primarily driven by the behavior of multi-license owners, they also benefit single-license owners, who as a group receive payout increases that are almost as large as those seen by multi-license owners. 

Limiting the distorting effects of strategic supply reduction

While the effects of strategic supply reduction are good for TV station owners, they may not be good for society as a whole. After all, the goal of the auction was to reallocate spectrum efficiently, rather than to generate profit for private equity firms.

So, can anything be done to limit the power of multi-license owners to sway payout prices in the reverse auction? 

We see two ways the design of the auction could be adjusted to take into consideration the joint ownership of TV stations more fully.

We propose two solutions. The first involves changing the rules of the auction for multi-license owners and recognizes that stations with higher broadcast volume are more important to the FCC in the repacking process. Under this proposed rule change, owners could not withhold a lower broadcast volume station they own from the auction without also withholding their higher broadcast volume stations. In other words, a station owner could not drive up the price on an essentially worthless station by strategically withholding a more valuable one.

We estimate that such a change would decrease the payouts from strategic bidding by somewhere between 71 and 89 percent. (Whether this rule change would survive potential legal challenges is more difficult to say.)

Another possible solution involves an adjustment to the repacking process. The FCC decided to take a conservative approach to the problem of interference, allowing only a .5 percent interference level—meaning that a TV station’s population served could not decrease by more than 0.5 percent because of the repacking process. Slightly increasing that figure, to 2 percent or even 10 percent, would have made it much easier for the FCC to reassign stations to new channels. As a result, withholding any one license from the auction would not have nearly as large an effect on payouts. 

Why strategic supply reduction matters

Some may ask why it is important to study a single auction this closely. We argue that this auction and its design had significant societal consequences, both positive and negative. 

In our paper, we identified strategic supply reduction as a theoretical—and very real—problem in the FCC’s incentive auction. Savvy players, we show, managed to exploit an assumption baked into the auction design, for their own gain and (arguably) the public’s detriment. As a result, the auction was less successful than it otherwise might have been and raised less money for taxpayers.

While we still see the auction as an overall success, we hope our work can contribute not only to the theoretical study of auction design but also to better real-world auctions in the future. 

This article summarizes ‘Ownership concentration and strategic supply reduction’ by Ulrich Doraszelski, Katja Seim, Michael Sinkinson, and Peichun Wang, published in the American Economic Review in March 2025.

Ulrich Doraszelski is at the Wharton School at the University of Pennsylvania. Katja Seim is at Yale University. Michael Sinkinson is at the Kellogg School of Management at Northwestern University. Peichun Wang is at the HKU Business School at The University of Hong Kong.