Antitrust & Competition Policy Blog

Editor: D. Daniel Sokol
University of Florida
Levin College of Law

A Member of the Law Professor Blogs Network

Tuesday, June 12, 2012

Asymetric switching costs can improve the predictive power of shy's model

Posted by D. Daniel Sokol

Evens Salies (Observatoire Francais des Conjonctures Economiques) discusses Asymetric switching costs can improve the predictive power of shy's model.

ABSTRACT: Economists Oz Shy introduced the definition of undercut-proof property (“UPP”) prices in a model of Bertrand competition involving loyal consumers (‘A quick-and-easy method for estimating switching costs’, International Journal of Industrial Organization, Vol. 20, pp. 71-87, 2002). Shy’s seminal paper allows applied researchers to measure the switching costs faced by locked-in consumers. Although there is increasing interest in demonstrating consumer inertia in retail markets opened up to competition, Shy’s approach has not received much attention. The present paper shows that the UPP’s lack of appeal in this context stems from a strong assumption of identical switching costs in the theoretical model, whereas real data are more likely to reveal asymmetric values for these costs. We revisit the UPP by considering asymmetric switching costs straight from the theoretical model. Doing so enables us to show that! more rigorous conditions relating the values of switching costs to market shares are necessary in order for UPP prices to be valid predictions of these costs, which consequently increases the predictive power of Shy’s model. This improvement is illustrated with two examples borrowed from Shy’s paper.

http://lawprofessors.typepad.com/antitrustprof_blog/2012/06/asymetric-switching-costs-can-improve-the-predictive-power-of-shys-model-.html

| Permalink

TrackBack URL for this entry:

http://www.typepad.com/services/trackback/6a00d8341bfae553ef01676681c0bc970b

Listed below are links to weblogs that reference Asymetric switching costs can improve the predictive power of shy's model :

Comments

Post a comment