[Thesis]. Manchester, UK: The University of Manchester; 2011.
A Rationale for Tying Merchants' Membership of Platforms Serving Independent Markets.
This thesis was submitted to the University of Manchester for the degree of Doctor
of Social Sciences in the Faculty of Humanities by Michael King during September 2010.
I analyses the effect of tying sellers' membership of a monopoly platform to membership
of another platform, which operates in an otherwise competitive market. Visa's contentious
use of the honour-all-cards rule to tie their debit and credit cards is an example
of such a tie-in. There has been a move to judge tying cases under "rule of reason",
which permits dubious practices when they are indispensable to creating economic benefit.
However, a proportion of the extra-surplus must be passed on to consumers ("pass on
test"). Rochet and Tirole (2008) claimed that tying payment cards raised Visa's
profit without harming end-users. However, this doesn't fully address the concerns
of regulators. Hence, my thesis investigates whether tying satisfied the "pass on
test". Part I: In Rochet-Tirole (2008) sellers operate in two independent markets
(├ž and d). Network A runs platforms in both markets; and Network B only operates in
market d. The price-level (buyer-fee plus seller-fee) on a network's platforms is
exogenously determined but they can choose the price-structure. My study extended
this framework by explicitly modelling competition in the product market. Part
II: Platform competition leads to a price-structure that maximizes the net-benefit
received by buyers and sellers. In contrast, a monopoly platform extracts most of
the surplus by encouraging excessive use of payment-cards. Therefore, if tying is
prohibited, then competition for sellers in market d leads to an optimal price-structure.
However, Network A extracts most of the surplus created by its monopoly platform.
Finally, if the average transaction-cost, ¤ä, exceeds the price-level, ¤ü, then the
net-benefit generated by a monopoly platform remains strictly positive. Part III:
By tying its platforms Network A can exclude Network B. However, Network A is unable
to exclude Network B just by matching the net-benefit it generates; rather, it must
"compensate" sellers for the extra competition they face from being on the same network.
Therefore, if tying is permitted, then the total net-benefit on Network A exceeds
the maximum benefit that can be generated by a single platform. Part IV: It was
found that if transaction-fees, ¤ü, are high relative to transaction-costs, ¤ä, then
tying always increases the consumer surplus. However, if transaction-fees, ¤ü, are
low relative to transaction-costs, ¤ä, then tying doesn't benefit consumers; and will
reduce the consumer-surplus if their transaction-costs are sufficiently high.