Spectrum Licence Renewals in Australia - Professor Richard Holden

69 existing spectrum licenses across 7 bands that are due to expire between 2028 and 2032. 48 of these are held by 3 mobile network operators and are of particular concern when it comes to the Australian taxpayer receiving appropriate compensation for the use of spectrum. Another important objective is ensuring that spectrum licenses are allocated efficiently.

This submission, written by Professor Richard Holden, and commissioned by ACCAN, responds to Australian Communications and Media Authority’s (ACMA) proposal regarding these Expiring Spectrum Licenses.

ACMA’s proposal is not consistent with its stated objective that “spectrum is a valuable public asset, and industry should pay a fair price for its use.”

The proposal makes a number of fundamental economic errors in reaching its conclusion. Implementation of the proposal would shortchange the Australian public, erode public trust in regulatory processes, and undermine competition in the telecommunications sector.

• The benchmarking analysis used to support a rollover of existing licenses at a reduced price to those historically paid is entirely without merit, based on a fundamentally flawed methodology, and provides no evidence in support of the conclusions drawn from it.
  

• The argument made for conservative pricing also has no basis whatsoever. There is no economic logic behind the claim that higher prices for spectrum licences lead to less investment in the network. Network investment will be made optimally by firms, trading off the marginal costs and marginal benefits of that investment. None of the firms in the Australian Wireless Broadband (WBB) spectrum context are capital constrained. The amount they pay licences is both sunk, and immaterial, at the point where they consider optimal network investments.

• The ACMA proposal will likely lead to an adverse impact on services to Australian consumers. The potential misallocation of licenses reduces competition, and may lead to a reduction in consumer choice and product offerings. Such effects include reduction in the probability of new entrants who would put downward pressure on pricing while expanding consumer choice, and constraining developments in the direct-to-handset market (enabled by low-earth-orbit satellites) which have significant growth prospects over the 16-year term of renewed licenses.

In contrast to ACMA’s proposal, this submission makes two core recommendations.

1. Recommendation 1: Expiring Spectrum Licences should be auctioned.

   To ensure that taxpayers receive an appropriate return for the use of their assets, to ensure that prices for those assets reflect relevant private information of bidders, and to ensure that the assets are allocated to the efficient user, an auction mechanism should be used.

2. Recommendation 2: International experts in market design should be consulted about the auction format, with particular attention to recent advances in combinatorial auctions. The auction mechanism and sale process would benefit from input from international experts with expertise in auction theory and the use of auctions in spectrum allocation. This includes the specific mechanism used, and the proposed grouping of different bands into different bundles.