So, it looks like I may have been a little unfair to Gruber in my last post. Or, at the least, a bit premature. I took advantage of the sun’s emergence this afternoon to sit outside and finish reading the final chapter in The Economics of Mobile Telecommunications. Though I had assumed that a final chapter before the Appendix would offer a conclusion and not a lot of further analysis, I found some very interesting observations and Gruber addressed many of the concerns I raised at the end of my last post.
The final chapter, ‘The evolution of market structure in mobile telecommunication markets’, also reinforced my acknowledgment that I’m not a natural economist. It took me about half an hour to really wrap my head around about 3 pages of material. To be fair to me, some of that was while in transit, which is not the easiest way to think about the implications of formulas like Π(n*, s, F) > 0 > Π(n* + 1, s, F) — the zero entry condition used to help determine the Cournot equilibrium number of firms in a homogeneous goods industry. [Note to self: need to read up more on industrial organization and oligopoly theory.]
Gruber used some theoretical models to assess actual 3G spectrum auctions in Europe, which were designed to improve competition in national telecommunication markets but may not have led to optimal outcomes. Sequential auctions in different countries resulted in decreasing licensing fees, as firms were better able to value licenses and evaluate international competitors strategies. Further, the earliest auctions were artificially-buoyed due to high telecom valuations and easy access to financing with the dotcom boom. One factor that I don’t think Gruber gave enough (any?) weight to was the UK and Germany being the 1st and 3rd auctions, which generated the two highest license fees per capita and, not so coincidentally, were the two strongest economies of auctioning countries.
One of the themes that comes across in the final chapter is that regulators tried to introduce pre-entry competition by automatically increasing the number of firms without much consideration of whether markets could truly support the additional companies. This was especially true with the auction process generating much higher then expected licensing fees and creating increased entry costs. The result, unsurprisingly with the move to a new and — at the time — emerging technology, was that regulators ended up needing to relax license network build out conditions as companies were unable to meet stipulated timelines. Yet, will governments struggling to balance national budgets be willing to forego that extra revenue in future auctions?
Gruber also notes that bidders may have over-estimated consumer demand for 3G services and concluded by stating,
The lesson to be drawn for the design of market structure is that the choice of the licence allocation mechanism has crucial importance for post-entry performance. The issue can be put starkly as that the regulator must determine whether there should be competition for the market or competition in the market. This may also require a rethink on the recourse to ‘market-based’ allocation mechanisms for public goods. Finally, in the light of the observed exit of firms one may also conclude that spectrum is no longer a scarce economic resource for the provision of mobile telecommunications services. These questions all provide scope for stimulating further research.
The first part of this conclusion rings true to me, especially for a country the size of Canada. We have the “Big 3” national carriers (Bell, Rogers, TELUS) and several regional/urban-centric operators (MTS, SaskTel, Vidéotron, Moblicity, Public Mobile, Wind Mobile, Shaw?) and a slew of independents and mobile virtual network operators. Even with Canada’s geographic size, there’s still only 34m people. Continuing to reserve spectrum for new entrants may not be the most effective for diffusion of new technology and driving down costs.
With exploding data consumption rates in advanced and rapidly developing countries, the last part of Gruber’s conclusion doesn’t stand the test of time. (The book was published in 2005, which was also the year that three former PayPal employees started a little video-hosting site called YouTube.) With the ability to download/upload HD content from mobile devices to the Internet, the spread of social networking and user-generated content, adoption of telepresence and the growth of cloud-based tools, regulators will need to continue to effectively manage spectrum allocation and governance.