Efficiency
and Robustnness of Ebay-like Online Feedback Mechanisms in Environments
With Moral Hazard
Chris
Dellarocas
Associate Professor of Management, MIT Sloan School
eBusiness
Research Seminars
NE20-336 (3 Cambridge Center)
Wednesday, February 26, 11:30 PM-1:00 PM
Refreshments Provided
Abstract
Online
reputation mechanisms are emerging as a promising approach for
building trust and inducing cooperation in online trading environments
where more established methods of social control (such as state
regulation or the threat of litigation) are often difficult or
too costly to implement. This paper offers a systematic exploration
of online reputation mechanism design issues in trading environments
with imperfect monitoring of a sellers effort level and
two possible qualities. The objective of reputation mechanisms
in such settings is to induce sellers to exert high effort. I
study a practically significant family of mechanisms that resemble
the one used by online auction house eBay. These eBay-like feedback
mechanisms solicit ratings of transaction outcomes as either positive
or negative and publish the sums of positive and negative ratings
posted by buyers on a seller during the N most recent periods.
I find that eBay-like mechanisms can induce high average levels
of cooperation that remain stable over time. Surprisingly, their
efficiency cannot be improved by summarizing larger numbers of
ratings or by publishing a sellers detailed feedback history.
They are quite robust to incomplete feedback submission and can
be made robust to sellers that can costlessly change identities
by setting the initial feedback profile of new players so that
it corresponds to the worst possible reputation. The
theoretical outcomes predicted by this paper are consistent with
empirical observations and offer new, theory-backed, explanations
to hitherto poorly understood phenomena such as the remarkably
low fraction of negative feedback on eBay. Finally, they provide
concrete suggestions on how eBays current mechanism can
be improved.
Calendar
for Spring 2003 Research seminars