A brilliant paper by Guru Acharya (don’t know the affiliation).
He shows how Lin Ostrom’s insights can be used to allocate spectrum.. It is a better strategy than auctioning to private players. The central idea is that spectrum should be seen as a common whose governance should be done accordingly:
This paper adopts Ostrom’s (1990) framework of governing common pool resources and applies it to electromagnetic spectrum in order to examine conventional policy prescriptions of privatisation and command-and-control, which are dominantly used in most jurisdictions for governance of spectrum. Additionally, Ostrom’s design principles for long-enduring, self-organised and self-governed common pool resources are used to propose a new institutional arrangement for governing spectrum.
The paper is structured in the following manner. First, Ostrom’s framework is introduced along with her critique of conventional policy prescriptions. Next, existing literature related to spectrum governance is reviewed. Finally, Ostrom’s framework is applied to electromagnetic spectrum as a common pool resource in order to examine existing prescriptions for governing spectrum and further to understand her prescriptions for governing spectrum.
Spectrum is a kind of common good – Rival (my consumption lowers your consumption) but non-excludable (difficult to price):
Traditionally, spectrum is considered as a classic example of a common pool resource characterised by high subtractability (rivalry) and high difficulty of exclusion (Savas, 1977). Subtractability of spectrum is a result of interference between electromagnetic waves, which are transmitted during wireless communications. Technically, interference is caused by superimposition of one electromagnetic wave onto another electromagnetic wave to form a resultant wave of greater or lower amplitude. Any undesirable electromagnetic wave which superimposes onto the intended wave is referred to as noise. Accordingly, since all wireless communications require transmission of electromagnetic waves, communication by one is noise for another. As a result, appropriation of spectrum by one person reduces its value for simultaneous appropriation by another person, thus making the resource rivalrous or subtractable.
The tragedy of commons is best described in this context by the phrases ‘chaos in the ether’ (The Milwaukee Sentinel, 1926) and ‘with everybody on the air, nobody could be heard’ (Nat. Broadcasting Co. v. US, 1943). However, the rivalrous nature of spectrum is a function of the technology deployed for transmission (Berge & Kranakis, 2011). In the recent past, many new technologies (like ultra-wide band) have emerged which have thepotential to push rivalry to almost negligible (Milgrom, Levin, & Eilat, 2011) but have not witnessed widespread deployment till date due to regulatory restrictions.
Both government interference and privatisaiton of spectrum are used to governs this commons:
For preventing the tragedy of commons in spectrum as a result of the metaphorical use of the three models discussed by Ostrom, the policy prescriptions dominantly used in most jurisdictions are the same as the conventional policy prescriptions critiqued by Ostrom (i.e. state intervention andprivate property). Of these two policy prescriptions for preventing the tragedy in spectrum, the most dominant prescription is that of state intervention in the form of command-and-control. The private property regime for spectrum governance is a recent phenomenon and finds its genesis in an article by Coase (1959).
In most jurisdictions, in order to coordinate wireless communications, the regulator divides spectrum into multiple frequency bands; assigns uses and technologies to different bands; and allocates bands to different (private) parties. Traditionally, the allocation to private parties has been in the form of lotteries, beauty contests or first-come-first-serve. Since the bands, uses, technologies and authorised users are administratively determined, this regulatory framework is referred to as command-and-control. By dividing spectrum into bands, each having specific uses, which are then allocated to certain entities, the state centralises control and administratively legitimises certain uses and users of the spectrum, thereby excluding the masses from using the spectrum.
In the recent two decades there has been a shift from the command-and-control paradigm towards the private property paradigm in which property rights are assigned to different spectrum bands (as a combination of time, space and frequency), which are auctioned in the market. Coase (1959), in his critique of the command-and-control mechanism, advocated that spectrum is a resource which can most efficiently be allocated using the market by assigning private property rights. The auction mechanism ensures that the private parties do not have any windfall gains, and the secondary market allows the spectrum to be reorganised efficiently according to size of band and the use for such band.
Applying Ostrom’s ideas to Spectrum:
Ostrom’s critique of external intervention is primarily on the issues that (i) external agencies do not have ufficient/accurate local information and that intervention based on inaccurate information can lead to lower payoffs; (ii) other costs like monitoring and enforcement costs can be relatively high for an external agency; and (iii) the focus of the external agency is only on a single arena.
If one were to use Ostrom’s critique of leviathan (external state intervention) and apply it to the command and control regime, various interesting points emerge. For example, use for spectrum bands is administratively defined by the center in the case of command-and-control management of spectrum. In this case, the central government may not have accurate information about localised demands (of use). Say, in a certain geographic area, more amount of spectrum is required for television broadcasting of local language channels than for mobile communication. Allocation that neglects such local requirements can lead to lower payoffs. Other similar examples also emerge. For example, the central government may not be aware that a particular area demands more spectrum for 4G services whereas the center may have created artificial shortage by allocating excessspectrum for 3G services.
Similarly, in an urban setting with high number of obstructions, low frequency does not have as good propagation characteristics as it otherwise does in rural areas creating further need for localised allocations. Similarly, interference, which is the primary reason for rivalry, is dependent on local topography, due to multipath interference, of which the central government does not have accurate knowledge.
The private property regime overcomes many of the shortcomings of the command-andcontrol regime. For example, by making the spectrum technology and service neutral at the time of privatisation, it allows the property owner to respond to market requirements. Therefore, the spectrum allocated can be used for providing any service using any technology as may be locally required. However, it still suffers from inherent limitations outlined in Ostrom’s critique. By dividing the spectrum into bands and privatising such bands, it suffers from similar defects as the field divided into small privately owned heterogeneous strips. Due to such privatisation, one operator may have spectrum in a low frequency band while another may have spectrum in a high frequency band. Or the same operator may have a high frequency band in one geographic area and a low frequency band in another geographic area. Worse, an operator may have spectrum in one geographical area but may not have any spectrum in another geographical area. This may lead to a number of problems/complications ..
So what to do?
Ostrom’s framework identifies two types of problems. The first is the appropriation problem and the second is the provision problem. Since spectrum is a perfectly renewable resource that is instantaneously replenished, the objective of the institutional arrangement reduces to solving the problem of appropriation only.
In the proposed institutional arrangement, a combination of spectrum in the high and low frequency bands is collectively allocated as a common property to a combination of multiple network service providers. The boundary of the resource system is the boundary of the allocated spectrum bands; and the boundary for the users is the collective of network service providers to whom the combined spectrum is allocated. All other users will be excluded from appropriating the resource. The proposed institutional arrangement requires these network service providers to then decide between themselves the appropriation rules. In contrast to the dedicated spectrum bands that are traditionally allocated under command-and-control and privatisation, in this case, the operators are required to collectively determine the rules for using/sharing the spectrum. It is expected that allocation rules will take into account the possibility of using technologies that reduce rivalry and will not merely involve dividing spectrum between member operators.
The same could be applied to India as well..
Keeping this brief background in context, if, for example, the government were to collectively allocate spectrum in the 900 MHz, 1800 MHz and 2100 MHz bands to a combination of 6 service providers at a pan India level and let the service providers mutually decide how to allocate spectrum between themselves, it would conform to the institutional arrangement proposed in this paper. In such an arrangement, all persons except for these 6 service providers would be excluded from using the allocated spectrum. These service providers could adopt flexible variations in the allocation rules to maximise their benefits. For instance, the service providers could collectively use the low frequency spectrum through active infrastructure sharing in rural areas to maximise coverage and minimise costs. Further, capacity constraints due to heavy usage in dense urban areas can be overcome by the shared spectrum which would have the combined capacity to serve all providers and would not lead to underutilisation in the hands of any one service provider.
Additionally, all service providers could have contiguous spectrum in all geographic areas permitting national roaming and nation wide competition. The service providers could mutually decide a harmonisation policy (across providers and across bands) to ensure that a healthy ecosystem of devices develops and economies of scale are best utilised. Lastly, service providers could also respond faster to market changes in a localised manner by providing/allocating more spectrum for voice based services in rural areas and more spectrum for data services in dense urban areas, which would otherwise be centrally decided by the Government.
Hmm.. How to identify the six? Via bidding again?
A very interesting idea and application though. The paper points it has been done in the past as well:
Wormbs (2011) discusses the design principles with reference to a historical institutional arrangement. Specifically, the paper discusses the case of frequency allocation in Europe in the 1920s in which the institutional arrangement conformed to the eight design principles enunciated by Ostrom. In this, European broadcasting companies self-organised under the aegis of the International Broadcasting Union and developed an institutional arrangement to prevent interference. This arrangement conformed to all of Ostrom’s design principles including boundary rules, allocation rules, monitoring and sanction.
So it has been done in the past as Most of Ostrom’s works are based on how the world really works and not how she wishes it would work. So if it was done in the past, there is some evidence. Need to read whether it was successful or not..
Very good paper indeed..