The idea of mobile operators jointly owning and operating the network is bubbling again. The latest was the call by the Banglalink CEO that all mobile operators cooperate in building one common 3G network (Growth hinges on telecom, Nov 11, 2010).
For an industry in which profits are under considerable pressure, the prospect of investing in an entire overlay network that may or may not yield revenues over and above what the existing 2G network currently generates is understandably scary. For governments that have been slow to re-farm frequencies, the prospect of getting 3G services with a smaller number of spectrum slots is attractive. For the many planners among us, still retaining elements of 1960s thinking, the idea of input optimisation is inherently attractive.
But the proposal is impractical and wrong. It should be scotched before it gains any traction in this key period where one hopes plans are being made for Bangladesh's 3G auctions.
Decision-making by consortium
Unlike traditional infrastructures such as water distribution systems -- telecom networks require continuous investment and improvement. Thirty years or even 60 years later, properly installed water pipes will function little differently than when they were put in. In contrast, the capabilities of a telecom network will be fundamentally different just a few years after it is built. Continuous software enhancements will increase the carrying capacity as well as functionalities.
This means that a telecom network requires a governance structure capable of quick investment decisions. It has been well established that government-owned enterprises lack this. What is less well known is that enterprises that are jointly owned by private entities that compete elsewhere on the value chain also lack the ability to take investment decisions.
An investment decision is not simply an expression of opinion; it has to be backed up by large sums of real money. Even in joint ventures made up of only two parties, it is difficult to make investment decisions. One may prefer the quick gratification of declaring dividends while the other is willing to delay the rewards in the belief that the delay will multiply them. The result is deadlock. Many joint ventures dissolve for this reason.
Can you imagine such decisions being taken by four or five operators with different profitability profiles and shareholders with varying expectations? A consortium of operators running a 3G network is a recipe for deadlock.
One may argue that the problem can be solved by majority vote. That is how it is solved in limited-liability companies or some partnerships. The rules decree that the minority must commit resources to implement the majority decision. But how realistic is this in the case of a consortium? Can a company that is losing money or is under pressure to keep its share price up, be compelled to invest? Unlikely.
The issue is not simply about upgrading the capabilities of networks. A 3G network is rarely, if ever, fully built out right at the start. Normal practice is to build in the main cities, the highways, then the secondary cities, and so on. The investment timing and extent depend on many factors, among them being the return on the previous investment. It is well known that different operators have different marketing strategies: some base theirs on the broadest possible coverage; other focus on different niches. They will obviously have differences on, for example, the pace of extending coverage beyond the highways into the interior to cover villages. These differences will create friction in all investment decisions taken by the consortium.
One company owns; others use
Even in the case of gas and water pipelines, where there is no continuous need for investment and upgrades, consortia are not optimal. Here, the best solution is seen as that which allows one company to build and operate the network, subject to an obligation to allow its competitors at the retail level to use it in a non-discriminatory manner and at cost-oriented prices.
Perhaps this is a solution for 3G networks, if an acceptable regulatory regime exists.
In the case of infrastructure that does not require continuous investment and upgrades, one may even embed the terms and conditions of giving access to competitors in a detailed concession contract. The government can craft this governing document using external expertise and transparent procedures. If the document is clear enough, it is possible that the need for ongoing supervision by a regulatory agency will be minimal. The problem, of course, is that if one party wants to renege on commitments or gain an advantage, all it has to do is hire clever lawyers. Then the need for adjudication and dispute resolution arises; a regulatory agency or arbitration becomes essential.
In the case of 3G networks, unambiguous concession contracts will be difficult to craft because markets and technologies are in rapid flux. Whatever is written down at a point of time cannot fully reflect all future trajectories.
That leaves the whole thing at the mercy of the regulatory agency and its discretion. If we are talking about building a 3G network in Sweden that will be made available to all operators on fair terms, I might concede the possibility of an effective regulatory solution. But the question is whether the single network can be made to work in Bangladesh, where there still are questions about the capacity and efficacy of the regulatory powers -- now divided between the Ministry and the Regulatory Commission.
The operation of multiple 3G networks, owned and operated by different limited-liability companies, may be sub-optimal if seen solely from resource optimisation or planning perspective. But it is actually the most efficacious for the ground conditions in Bangladesh.
The different networks will be rolled out at different speeds, and be based on different geographical and service-quality priorities. Consumers will have real choices and things will get done with minimal dependence on high performance by the regulatory agency. We go back to a fundamental truth: the best solution depends on the fit with actual circumstances, not fealty to an abstract ideal.