ARCEP Conference on Mobile Economics Paris, 26 March 2007 ‘Should mobile termination rates differ?’ Martin Cave. Warwick Business School, UK [email protected] Issues . When should an NRA set differential rates for mobile operators? Should they converge? If so, over what period?
Commission comments to NRAs on symmetry-eg
-period of asymmetry must be justified, ‘based on a cost model which takes into account costs of an efficient operator and the complete process of adequate accounting information to be provided by all MNOs’
-NRA is invited to monitor the cost structures and assess whether its current assumptions on ‘fair and reasonable prices’ will remain relevant.
Four possible regulatory objectives(some ambiguous)
not a good basis for differential rates!
- consistent with competitive outcome and
with efficiency-enhancing prices
- incentive problems under CPP
i) with competitive spectrum markets, or
‘opportunity cost’ prices: no problem as prices take the strain
ii) with arbitrary fees or historic valuations: allow cost recovery or promote efficiency/competition
D Location of traffic
-Should geographically-based termination cost differences be incorporated, as detemined by cost function?
-Impact depends on pass-through to retail prices
-Call-by-call differentiation not feasible, but operator-by-operator differentiation promotes cost orientation and competitive parity
G-H Different start dates/market shares
Dynamic efficiency argument: competitive advantage required for late-comers for sake of long-run end-user benefits
Precedents include OPTA’s ‘delayed reciprocity’ in fixed termination rates and Ofcom’s WBA margin squeeze restrictions
Justification should relate to effective start date in the first instance, not market share
Factors which influence the decision include:
-entry date gap and time elapsed
-maturity of market
-expected technological and regulatory developments
-impact of traffic asymmetries
Evidence that ‘slow’ mobile number portability (MNP) has a weakened/ insignificant effect on churn (Lyons 2006), whereas
‘fast’ MNP lowers prices and increases churn
Combined with contract length, this permits calculation of period of effective switching needed to neutralise earlier entry
Relative size does not skew inter-operator termination cash flows where traffic is balanced
Where it is unbalanced, excessive termination rates disadvantage operator with a trade ‘deficit’
Combined with low on-net prices charged by ‘surplus’ operators, this can tip deficit operators out of the market (and may even breach competition law)
Is this a basis for differentiation?
Some grounds for differentiation are unexceptionable
-higher costs from geographical factors
-time-of day traffic differences
Others are more debatable
-coexistence of different networks
Problems arise with inclusion of dynamic factors
- regime 1:differentiation can sustain inefficient operators dependent on regulatory favours, which do not benefit end-users, OR
-regime 2:differentiation for a short period may prevent numbers in the long-term going down from 3 to 2 or 4 to 3.
Possible solution lies in estimating realistic period of ‘hope’ for regime 2 to operate, and committing and sticking to differentiated rates acordingly.