Late last year I had the pleasure of participating in an event on Spectrum Auctions put together by the New America Foundation.  I’ve blogged about the event before, but when it was recently reported that T-Mobile’s CFO, Braxton Carter, stated that consolidation in the mobile wireless sector was inevitable (“It’s not a question of if, it is a question of when”), I was reminded of an interesting anecdote provided by one of the event’s other participants—former FCC Chairman Reed Hundt.    

Specifically, Chairman Hundt was recounting his experience in designing and implementing the first PCS spectrum auctions back in the 1990s.  (Watch the first video here, beginning at about 25.50.) Given the lack of precedent, Congress tasked the FCC with the difficult job of having to figure out both the big picture and small details of the auction process, including how many licenses to auction in any given market.  This number would, in part, determine the number of firms offering mobile wireless services (in addition to the two already in operation).  On this question, Chairman Hundt indicated he relied heavily on the advice of Professor Michael Porter at the Harvard Business School.  The government, the Professor rightly reasoned, could not pick the equilibrium number of firms in advance.  So, Professor Porter recommended that the FCC should sell more licenses than it believed the market would bear.  By doing so, the equilibrium number of firms in the market could be discovered since, eventually, the extra firms will go bust and another firm will buy them.  What would happen over time, Chairman Hundt indicated, is that the many would consolidate to the few as the market matured.  (If you’re wondering about this idea of equilibrium number of firms, read our paper Competition After Unbundling:  Entry, Industry Structure, and Convergence.) 

To sum up, the advice from Professor Porter was to sell too many licenses and let consolidation move the industry to its equilibrium.  And that’s what the FCC did, providing a shot for up to seven firms to offer mobile wireless services.  

In light of Chairman Hundt’s “overshoot the equilibrium plan,” I find the Obama Administration’s hard line and much of the current debate over the price-increasing effects of further consolidation misplaced.  This “start with too many” approach and let consolidation do the rest is based on basic economic logic.  That logic holds that when there are too many firms, firms lose money (some or all of them).  Consolidation may weaken price competition and/or improve efficiency, and it also divides the industry output among fewer players, thereby increasing profits so that the remaining firms have financially viable operations.  Consolidation will continue until all firms are viable, and at this point, the equilibrium is reached. 

In this strategy devised by Professor Porter and Chairman Hundt, even if we limit the argument solely to a price effect, permitting consolidation from a point known to be “too many” can still be good policy.  In fact, it is essential, since eventually some firms will simply go away due to financial losses.  Consolidation can occur via merger or via exit.  (Most firms and regulators prefer acquisition to exit, since the former maintains value and eases customer transition.)  In effect, given the way in which the FCC rigged the game, higher prices from a merger is not a valid argument against a merger.  Higher prices are a direct and arguably planned consequence of the government’s strategy with regard to spectrum auctions and the subsequent groping toward equilibrium industry structure.

Consider a simple numerical example based on the Cournot Model of Competition (details of the basic setup are outlined here).  Let market expenditures be $90, the per-firm fixed/sunk cost of entry be $10, marginal cost be $1, and all firms identical.  Table 1 summarizes the resulting industry price, each firm’s output, and each firm’s profit at different numbers of competitors.  Under these conditions, the equilibrium number of firms (N*) is 3.  If there are 6 rivals, then all firms lose $7.50, which is unsustainable.  The exit of one firm leaves 5 rivals and a higher price ($1.20 to $1.25), but profits are still negative at -$6.40.  Another firm exits, leaving 4, and price rises to $1.33, but still profits are negative (-4.38).  Once the 4^th firm exits and price rises to $1.50, the remaining three firms breakeven with zero profits.  The equilibrium is reached.  (Note that profits need not be zero, just small enough so that 4^th firm won’t fit into the market.)  This simple example is an illustration of the plan chosen by the FCC with regard to spectrum and the mobile wireless industry.  Here, I have assumed there’s no efficiency from consolidation or other technological advancements that lead to lower prices, but such considerations were not necessarily part of the FCC’s plan when implemented.

Has consolidation in the wireless industry increased prices?  While there are predictions of dire consequences at every merger announcement, there’s no evidence that consolidation has driven up prices.  In fact, average revenue per minute has fallen consistently over time as consolidation has risen.  (Notably, it’s not clear what average revenue per minute means in the modern, data-centric market.)  Moreover, with spectrum exhaust, we have demonstrated that there’s good reason to expect that consolidation will lower industry prices (and not because of the threat of a price war).

Chairman Hundt also said that Professor Porter predicted that in about 20 years it would be necessary for “real political and thought based-courage to be shown [because] there would be a point at which the government would have to say that the consolidation can’t go on any further.”  What was unclear to me is whether Chairman Hundt meant that this courage would halt consolidation before or after the equilibrium industry structure was reached.  In either case, the critical question now is whether that equilibrium is three or four “nationwide” providers.  (There are in fact many regional providers and multiple firms selling attractively-priced nationwide services as MVNOs, but let’s stay focused on the debate as it presently manifests.)  Thus far, the government is betting on four (and recent press reports indicate that the DOJ has no indication of changing its mind anytime soon); the industry, it appears, thinks it’s three. 

Interestingly, the “Rule of Three” appears the norm, not the exception.  On that, I direct you to a global review of industry structure provided by Chetan Sharma (Competition and the Evolution of Mobile Markets).  In that paper, Chetan concludes,

Of the 40 major markets we studied, on average the top 3 mobile operators controlled 93% of their respective markets.  In fact, 36 of the 40 markets had the top 3 players control 85% or more of the market.  Other hyper-competitive markets like the UK and the US, which had more than 4-5 large players are moving towards the consolidation phase wherein the top 3 control more than 80% of the market.  Of all the large major markets, only India proves to be an exception and it is due to the hyper-growth phase of the market and due to somewhat unique nature of the market (p. 7).

Additional evidence supporting the “Rule of Three” is provided in the FCC’s 16th Annual CRMS Report.  (See Table 2 below).  Given the Political Interest Groups’ arguments that the U.S. ranks poorly in global comparisons (see, e.g., the New America Foundation’s flawed Cost of Connectivity Reports), it appears they should be happy to see further consolidation so that the U.S. industry structure better comports with its global counterparts.  Such groups have labeled the U.S., the U.K., and Canada as “fall[ing] in the high to middle price tiers” with countries like Sweden and Finland having “relatively cheap rates.”  Based on Table 2, these groups should be pleased to see further consolidation.  (But, see here and here for conflicting evidence.)

Beginning with the first spectrum auction, the government has admitted ignorance on market structure and thus bet on market-driven consolidation to reach industry equilibrium.  We now face the difficult question of whether the equilibrium it is three or is it four national firms (roughly speaking, since there are regional providers and MVNO’s that add to the competitive dynamic of the industry).  So while Chairman Hundt hopes for the “real political and thought based-courage” to halt consolidation, perhaps what is needed is the courage to commit to the plan he implemented and let the market find the market equilibrium.  Since the DOJ can’t count to less than four or think of any outcome but price, good merger policy in the mobile wireless sector must come from a courageous and expert FCC—one strong and wise enough to tell the DOJ to pound sand.  It’s happened before (see U.S. v. FCC).

Is the equilibrium three or four national firms?  I pass no judgment today.  But, given the setup of Chairman Hundt’s FCC two-decades ago, getting an answer to that critical question will require honest, rigorous, and dispassionate analysis of industry structure.  Significantly, under the Hundt-Porter strategy, even if we know a merger will lead to higher prices, this knowledge is not a sufficient reason to block a merger.  Higher prices was, in large part, the plan.