With termination rates fixed to costs the change in demand is:
∂ ∑j qj,i
∂ki
∂ Pj qi,j
∂ki
-c0i(ki)si j sjbj
-c0i(ki)si (1 - si)bi
(10)
The demand effect is twice the demand effect with linear prices. With the additional
assumption of termination rates equal to per-minute costs it is four times the effect.
This excess demand increase does not change j ’s off-net profit as retail prices are at
per-unit costs.
So far I have ignored the impact on the subscription fee. With a lower per-unit price the
subscription fee is set higher dependent on the (expected) increase in consumer surplus.
As Peitz (2005) states:
"In a neighborhood around cost-based access prices an increase in the competitor’s
[entrant’s] access price leads to lower subscription fees of any [both] operators."6
Taking for example the model of Peitz (2005) one can easily show that the investor’s
and a competitor’s subscription fee rise due to investments. Corresponding results could
be derived for the investment effect on own profits from outgoing calls and also for an
indirect investment effect on outgoing calls.
In praxis the traffic-independent subscription fee of two-part tariff schemes cannot be
adjusted as fast as the per-unit price. The adjustment delay is due to contract duration
with customers and overlapping beginnings of contract periods. As investments do not
have a one-shot property but, moreover, are conducted continuously one should either
expect ongoing re-adjustments of subscription fees or, alternatively, per-unit prices above
per-unit costs as a second-best option. Contract adjustments are not implausible even
during the contract period if providers benefit from lower retail prices, for example due
to own or competitors’ investments.
3.4 Regulation
With the transition of the competition enforcing regulation directives to national law in
2004 and 2005 mobile termination markets have been regulated in all EU member states
at the latest. Nevertheless, the directives do not specify which regulation scheme should
be adopted. In European countries mainly two alternative regulation schemes have been
in place since 2000. Cost-based regulation like LR(A)IC is the most wide-spread ap-
6 Peitz (2005), p. 9. Peitz considers the two-provider-case.
10