Regulator flexibility and the administrative allocation licensing of 3G spectrum

Abstract

Globally, most nations assign radio spectrum to provide 3G mobile services during the period 1999–2007. While there is consensus among most economists that auctions are the preferred assignment mechanism, the assignment mode is split, more or less, equally (in terms of the number of licences issued) between administrative allocations and auctions. With auction procedures tending to raise more revenue for governments (Cartelier, 2003) the question that naturally arises is: why are administrative allocations so popular a method to assign spectrum? McMillan (1995) conjectures that administrative allocations provide additional ‘flexibility’. Accordingly, this study examines the performance of 3G assignments in terms of an econometric analysis of a unique sample of national 3G spectrum administrative allocations. These outcomes are modelled as depending on spectrum package attributes, and post-award network deployment requirements.

Keywords:

• wireless telephone markets
• 3G spectrum assignment
• administrative allocations

JEL Classification::

• D82
• L51
• L96

Acknowledgements

The MIT Center for Digital Business and the Columbia Institute for Tele-Information provided support during the revision of this article. Helpful comments were provided by participants in seminars at Columbia University and Curtin University. The authors are responsible for all remaining errors.

Notes

¹ For example, under the EU regulatory framework, it is the responsibility of Member States to determine the mechanism for spectrum licensing and to define the licensing conditions applicable to its territory (EC, 2001). Member States must follow certain principles laid down in EU legislation requiring that licences be granted through open, nondiscriminatory and transparent procedures, based on objective criteria defined in advance.

² The stylized description of an administrative allocation for the purpose of this study is: firms make an offer that specifies service levels, quality, implementation speed and price. Offers are discussed with the NRA, compared to other offers, and modified by participating firms. Based on final offers NRAs decide the winning firms.

³ Greater market competition and diffusion of mobile technology can also help facilitate improvements in other potential NRA objectives not outlined in tender documents such as higher industry productivity (Ng, 2011) and service quality (Resende and Tupper, 2009).

⁴ Whilst infrastructure sharing may increase competition in the short-run, the development of facilities-based competition is compromised. Thus, NRAs face a trade-off between short- and long-run competition goals through the choice of sharing arrangement specified in tender documents.

⁵ The ‘cluster’ estimator is intended to ‘correct’ estimated SEs for panel data sorts of effects that are present, but omitted from the model.