The MLI paper on telecommunications policy




Chris MacDonald, lately a commissioner of the CRTC, and Peter Menzies, former Vice Chairman for Telecommunications at the same agency, recently published a paper called Building Internet Access is Job 1.We think it is a useful contribution to public discussion of matters of national importance. There is much that the Internet Society likes. Before we get on to the subject of MVNOs[1], where we differ, we want to set forth our agreements of several large issues.


Its central thesis is that regulation involves bargains where various conflicting objectives must be balanced.


“Decision-makers today face a balancing act. Fostering infrastructure investment and reducing prices for consumers are competing objectives. Telecoms is a capital-intensive industry, and the cost of connectivity and service in rural and remote Canada is exorbitant. Therefore, a decision adding significant cost to industry may negatively affect investment in new infrastructure, particularly in rural areas. But the people who live and work in those remote areas are citizens too, and must have quality service at an affordable price to be able to participate in the economy and digital society.”




The paper has the great merit of seeing the subject of broadcasting as a topic of narrow concern, relative to access to and the affordability of the Internet. Indeed it uses the word “distraction” in relation to the BTLR Report and its special pleading for journalists and broadcasters. Second, MacDonald and Menzies are firm in support of net neutrality, which is the obligation of carriers, within reasonable limits imposed by the needs for appropriate network management, to carry traffic without undue discriminations.


The MLI report makes a few recommendations on which we are not so keen. For instance, the ISCC is concerned that the regulator not become a policeman, and so we are skeptical of administrative monetary penalties (AMPs), both on grounds of natural justice and because the power to impose them can generate a conflict of interest within the regulator.


However, ISCC does not wish to be distracted from our main concern, which is the subject of MVNOs. The MLI paper sees the issue in terms of the balance between the need for investment, which the authors take to be the paramount concern, and the need for lower consumer prices. Though they never explicitly say so, the desirability of lower consumer prices is sacrificed on the altar of funds in the hands of carriers which are to be dedicated to infrastructure investments.




We think this opposition is mistaken in fact and the desired outcome bound to be frustrated. We will try to explain why.


The emphasis of regulatory bargains at the beginning of the paper directs our attention to what the authors are trying to tell us, namely that you can have lower consumer prices or greater investment in rural broadband deployment, but not both. Implicitly the argument is advanced that the big carriers will be efficient transfer mechanisms of revenues from consumers into rural and remote investments.  Allowing MVNOs will generate costs for the industry, in the form of foregone profits. So that, by reducing the profits available to the larger carriers, Canada could end up depriving people of the investments they need to gain adequate internet access.


This argument has powerful force and powerful backers. Nevertheless, we are unpersuaded, for several reasons.


The authors make four recommendations regarding mandated MVNO access:


To minimize the potential risk, any decision to mandate MVNO access should:


  1. Be a decision of the CRTC, not the federal government.


  1. Be limited in scope to minimize the potential negative impacts on continued investment by facilities-based carriers.


  1. Be organized in a manner that supports the rapid expansion of facilities-based carriers that own spectrum and have deployed regional wireless networks.


  1. Have an established end date.


First, the premise of the argument. If the price of access is set appropriately, carriers ought to be fully compensated for their costs in supplying access to intermediaries such as MVNOs, resellers and the like. Putting it another way, the MVNO must supply a customer to the underlying carrier at less cost than the carrier can do for itself, or else the MVNO goes out of business.






Second, MVNOs have every incentive to generate efficiency through innovation in computer technology and business processes. Their input price is fixed by the rates for wholesale access which have been set by the Commission. Their only way to compete with big carriers is through improved computing technology, better service or both. Such incentives do not exist for the big carriers or do not exist until they face increased competition. Thus MVNOs will present a constant source of competition by innovation, which can only benefit the ultimate user, the consumer.




The authors proceed from the premise that owning facilities is essential to be a successful carrier. This is like saying you cannot be in the car leasing business unless you produce cars.


The negative impact on investments is also questionable. After all the large carriers will sell capacity to MVNOs at a price matching costs plus a profit. In effect the MVNOs constitute guaranteed customers which require no marketing or servicing and who generate a predictable profit.


The authors clearly believe that competition can only come from regional carriers who own facilities and spectrum. If regional carriers have spectrum and facilities, they are not MVNOs. In addition, the assumption that competition can only come from regional carriers who own facilities and spectrum is an assumption for which there is no evidence. Indeed the many and repeated failures of facilities-based entrants in Canada argue strongly that this latest attempt to generate facilities based competition will fail, as previous rounds of new licences have failed.


Finally, if there is an end date for MVNOs one asks why anyone would get into the MVNO game at all, if the choice is either to grow into a facilities-based carrier or to sell out. Selling out under the pressure of an approaching end date clearly will have a depressing effect on the sale price.


In short, the basic assumption of the authors that competition can only come from regional carriers who own facilities and spectrum leads them to very faulty conclusions regarding MVNOs. 




But the argument is not just about the desirability of facilities-based carriers, rather than MVNOs. The basic objection to MVNOs is that they lower the profits of the larger incumbent carriers and that this should not be done.




The authors assume that the very existence of MVNOs lowers carrier profits. Why would this happen if the MVNO is supplying the customer and still making a profit for itself? That would imply that the underlying carrier is making supernormal profits, profits that would not be available but for its market power. Market power is the condition in which you can extract more profit from an activity than would be available under effective competition.


That leads us to the core of the argument for the carrier position. If MVNOs reduce the market power of the large carriers, there will be less money for redistribution to rural broadband, say, or any other social objective.


To our eyes, the argument against MVNOs comes down to this: leave the market power of the underlying carriers alone. They will use this surplus money to invest in rural broadband and other potentially unprofitable or less profitable social objectives.


The “cost” added to the carrier industry by the existence of MVNOs turns out to be reduced profits. Their market power is reduced by an effective policy intervention.  Remember that they are being paid – by regulatory decision – a price to the MVNO for access that meets all their costs and adds a measure of profit.




This is apparently not enough. Hence the need to talk about “regulatory bargains”, which is a kind of euphemism for not addressing the issue of market power.




As MacDonald and Menzies explain:


The cost of connectivity and service in rural and remote Canada is exorbitant, and the return on investment will always be such that it makes more economic sense for a shareholder value-based company to build a cell tower in Etobicoke than in Fort Providence. There will always be a better business case for launching a service-based competitor in Burlington than in Baker Lake.


All true. But leaving supernormal profits in the hands of the larger carriers is unlikely to lead to the realization of the goals that MacDonald and Menzies ardently desire, in our opinion. While they speak of each decision of the CRTC being predicated on market forces as much as possible, in practice they reject the one proven effective force for increasing competition, which has been shown to be resale of underlying capacity.




This brings us to the second major section of their argument, which is that the profits left in the large carriers will somehow be translated into investment in less profitable or unprofitable rural and remote broadband access.


To which we ask: why? Why would any rational capitalist organization reduce its return on investment to satisfy the distributive goals of the Canadian state? In various times the management of Canadian carriers have declared that maintaining investor confidence through high and continuing profits is the basis of their ability to attract capital.


We see no reason to believe that higher profits go anywhere than to investors. Absent a degree of coercion that a government would be reluctant to use, profits will be directed to investors, as they are expected to be.


Apart from these arguments in principle we also observe that MVNOs are normal. Today competition via MVNOs is offered in 79 countries. Wikipedia reports:


As of December 2018, there were 1,300 active MVNOs operating in 79 countries, representing more than 220 million mobile connections – or approximately 2.46% of the total 8.9 billion mobile connections in the world. The eight countries with the largest number of active MVNOs in 2018 were: USA with 139 MVNOs (4.7% market share), Germany with 135 (19.5% market share), Japan 83 (10.6%), UK 77 (15.9%), Australia 66 (13.1%), Spain 63 (11.5%), France 53 (11.2%) and Denmark 49 (34.6% market share).[16]




In my time at the Commission (2009-2013) and since, I have not seen any published recognition by the CRTC, or a study, which has examined the world-wide phenomenon of MVNOs. Our thinking on this subject remains parochial.


It should be noted that not all of these countries require access to underlying facilities by regulatory decision. However, most governments and regulatory agencies that have allowed MVNOs have also taken steps to regulate the terms on which MVNOs gain access to underlying carriers.


We have to ask ourselves why Canada is so reluctant to allow MVNOs and to regulate the terms on which they might access the carriers?


At the back of many minds is the unspoken assumption that competition made feasible by regulatory arrangements is not real competition. This is to ignore the extent the market power of incumbents is the creation of government policy in times past, and the continuing effect of licences and legal privileges that persist. New regulatory arrangements are visible acts of policy, old ones look like ivy-covered buildings from time immemorial, when they are merely 19th and early 20th century compromises.




There is a strain of thought, seldom expressed but assumed, that facilities-based competition, the kind where you own the network end to end, is somehow better, more real, than competition based on regulatory arrangements. This view is assiduously marketed by many “free market” think tanks.


This idea, deeply believed by those paid to believe it, and persuasive to many who have not thought about it at length, is antithetical to the capitalist idea, which as Hayek observed, is a discovery process. It presumes that there is only one method to get to a competitive result. In fact, no one knows the optimal combination of materials, methods, and processes whereby competition is to be achieved.  And if they are known for one period of time, they change in situations of rapid technological advance. Software can substitute for hardware. Virtual paths can substitute for wires. The facilities-based obsession is a kind of throwback to central planning carried out by governments, at the urge of carriers who benefit from the arrangement.


The other strains of argument of against MVNOs are the density of population, or the cold, or anything. Countries as vast and sparsely populated as Australia have mandated MVNOs. Northern Sweden, Norway and Finland have condition as cold, and MVNOs. There is always some excuse for thinking Canada special and unique. A thorough survey of the 79 countries that have MVNOs might open some minds, but so far we have not seen it undertaken by Industry Canada or the CRTC.




Oh, and then there is 5G. 5G relies on ultrashort frequencies and therefore small cell sizes, requiring lots of new equipment to serve the microcells. Hence the carriers need to be protected from MVNO competition because 5G investments need, you guessed it, higher profits that competition would undermine.




There is never a lack of reasons why carrier market power needs protection: national sovereignty, population dispersion, the cold, the heat, the size of the country, 5G. Anything will do, as long as policy makers believe it. In movies they call this the “MacGuffin”, the thing around which the plot is structured: the Maltese falcon, the missing nuclear codes, the suitcase full of cash. In telecommunications, think 5G, and population density.




Higher than normal profits is the real objective. Suppressing competition is the actual means to achieve it. Anything that suppresses an inclination to more effective competition will suffice. Alas, there is never a lack of policy makers who are not in the regulatory game for long enough,  who turn their eyes from foreign examples of success, or who persist in their sincere belief that to lower incumbent carrier profits is to engage in bad policy.




And in that regard, the MLI paper is another case in point.


Too bad, because MacDonald and Menzies are sound on many issues of national importance, about which we have here not said enough. But on the issue of MVNOs, they have erred greatly.





I iwsh to thank Konrad von Finckenstein for his comments, which I incorporated.




















[1] Mobile Virtual Network Operators, or MVNOs, buy bulk capacity from larger carriers and resell that capacity, sometimes adding or changing the nature of the underlying service, or marketing it in a different way or to different sets of consumers.

To top