Title: Economic theory of intercarrier compensation: twosided markets Tommaso Valletti Imperial College Lon
1Economic theory of intercarrier compensation
two-sided marketsTommaso VallettiImperial
College London and CEPR
2Two-sided markets and interconnection
- Why subscribe to a network? Ability to exchange
messages between parties. - Consumption involves both a sender and a
receiver. - Networks are platforms that bring together
senders and receivers. - Users typically subscribe to different networks.
3Two-sided markets and interconnection
- There are externalities (network call).
- Prices play a role in internalising
externalities. - Crucial role of interconnection intercarrier
compensation rules affect efficiency and market
structure. - Two roles (at least) source of revenue affect
the structure of traffic-sensitive retail prices.
4Who should pay for interconnection?
- Interconnection is costly. Cost causation.
- Does the sender cause costs by making a call? If
so then the terminating network should recover
from the sender (directly or indirectly). - Does the receiver cause costs by accepting a
call? Then the other way around! - Originating network buys termination (access),
or terminating network buys origination? - Mere conventions. Outgoing and incoming services
are complementary.
5Literature on two-way access pricing
- Earlier literature (Armstrong,
Laffont-Rey-Tirole) - Access charges affect retail pricing and users
consumption decisions. - collusive nature of access charges YES with
simple (e.g., linear) pricing structures, NO with
more sophisticated (e.g., two parts). - Overall, not a big a concern for policy makers
compared to one-way interconnection
(foreclosure). - Allow for reciprocal negotiations if possible
(applies to M2M, not F2M). - One crucial missing bit utility only from making
calls, no interaction sender/receiver (no call
externality). - Only sender causes cost access charge
recovered by terminating operator cost of
termination (LRIC).
6More recent studies with receivers benefits
- De Graba
- Jeon-Laffont-Tirole
- Hermalin-Katz
- Cambini-Valletti
7Efficient pricing structures
- When sender benefits are a fixed proportion of
total benefits from exchange efficient network
utilisation attained when total cost allocated to
sending and receiving parties in proportion to
benefits received (DeGraba). - Outgoing and incoming prices sum to marginal
cost. - With CPP, sender is king. Outgoing price should
typically be below total cost to generate more
calls.
8Unregulated (competitive) pricing
- Possible problem, connectivity breakdown using
off-net discriminatory pricing (J-L-T) - If users care a lot about receiving calls -gt set
very high off-net price (punishe the rival
relatively more). - If users care a lot about sending calls -gt set
very high reception charge (under RPP) which
destroys the rivals ability to make outgoing
calls. - But, this problem much reduced in the presence of
call propagation (call-back, Cambini-Valletti). - Empirical evidence (1 call sent -gt at least ½
calls received).
9Role of access charges
- When access below cost break-down problem much
reduced (off-net price pushed down). - Firms themselves have an interest not to induce
break-down (e.g., by choosing low access
charges) with b-d competition for customers is
toughest! - (Hold if no tipping effects.)
- If traffic is unbalanced targeted entry
(Hermalin-Katz). Positive access charges induce
entrants to target mostly users who receive calls
but do not originate many -gt target fewer end
users -gt soften competition.
10Bill keep
- Usage. Access charges affect efficiency via
usage-sensitive retail prices. BK optimal when
senders and receivers equally benefit. - Connectivity breakdown. Reduced risk.
- Investments. BK increases investment! (Think
about opposite, i.e., a gt LRIC. If network jumps
ahead of a rival -gt traffic imbalances -gt access
deficit! -gt No incentive to invest). - Targeted entry. BK prevents strategic behaviour.
Possible divergence between total welfare and
consumer surplus. If entry needed to bring prices
down -gt may accept even inefficient entrants -gt
attracted only if can make targeted offers.
11Mobile networks
12RPP vs CPP
- RPP vs CPP.
- Empty debate if we do not say how access charges
are set. - RPP already doable in Europe (intl roaming). At
present, F2M termination well above cost -gt no
reason to charge own customers for receiving
calls. - BK might induce RPP without mandating it.
13RPP vs CPP
14Mobile-to-mobile vs fixed-to-mobile
- M2M vs F2M
- Very different economics
- Strategic interaction for M2M/(largely) no
interaction for F2M - Possible bargaining for M2M/no bargaining for F2M
(asymmetric treatment of termination on fixed) - Allow for differential treatment of the two sets
of access charges (undone if no discrimination
requirement). - M2M BK should emerge!
15On-net/off-net price discrimination
- Typical complaint on-net prices of the large
network are lower than its termination charge. - Pon price on net, Poff price off net
- co origination cost, ct termination cost
- a (reciprocal) access charge (termination).
- A prohibition to have a higher margin off net
(prevent predation) - (1) (Pon co ct) (Poff co a)
- A prohibition to discriminate over incoming calls
(prevent foreclosure) - (Pon co ct) (a ct)
16On-net/off-net price discrimination
- co and ct are not easy to observe -gt set them to
zero (a conservative assumption) - (1bis) a Poff Pon,
- (2bis) a Pon.
- Typical complaint could mean principally that
either on-net prices are below cost, or that the
termination charge is far above cost. - Important role played by access charges.
17- Danke schön! - Thank you!
- t.valletti_at_imperial.ac.uk
- http//www.imperial.ac.uk/people/t.valletti