When Do Markets Tip? A Cognitive Hierarchy Approach

Posted: 13 Nov 2013

See all articles by Tanjim Hossain

Tanjim Hossain

University of Toronto

John Morgan

University of California, Berkeley - Economic Analysis & Policy Group

Abstract

The market structure of platform competition is critically important to managers and policy makers. Network effects in these markets predict concentrated industry structures, whereas competitive effects and differentiation suggest the opposite. Standard theory offers little guidance — full rationality models have multiple equilibria with wildly varying market concentration. We relax full rationality in favor of a boundedly rational cognitive hierarchy model. Even small departures from full rationality allow sharp predictions — there is a unique equilibrium in every case. When participants single-home and platforms are vertically differentiated, a single dominant platform emerges. Multihoming can give rise to a strong–weak market structure: one platform is accessed by all, and the other is used as a backup by some agents. Horizontal differentiation, in contrast, leads to fragmentation. Differentiation, rather than competitive effects, mainly determines market structure.

Keywords: platform competition, bounded rationality, cognitive hierarchy, vertical and horizontal differentiation

Suggested Citation

Hossain, Tanjim and Morgan, John, When Do Markets Tip? A Cognitive Hierarchy Approach. Marketing Science, Vol. 32, No. 3, 2013; pp. 431-453; DOI: 10.1287/mksc.1120.0770, Available at SSRN: https://ssrn.com/abstract=2329014

Tanjim Hossain (Contact Author)

University of Toronto ( email )

105 St George Street
Toronto, Ontario M5S 3G8
Canada

John Morgan

University of California, Berkeley - Economic Analysis & Policy Group ( email )

Berkeley, CA 94720
United States
510-642-2669 (Phone)
810-885-5959 (Fax)

HOME PAGE: http://faculty.haas.berkeley.edu/rjmorgan/

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
511
PlumX Metrics