Optimal Design of Multi-Stage Elimination Tournaments with Heterogeneous Agents: Theory and Experimental Evidence

*Uwe Sunde
Economics Department, University of St.Gallen

Rudi Stracke
Economics Department, University of St.Gallen

Rudi Kerschbamer
University of Innsbruck

Wolfgang Hoechtl
University of Innsbruck

     Full text: PDF
     Last modified: December 1, 2010
     Presentation date: 03/13/2011 12:00 PM in NH 1120, Session A
     (View Schedule)

Abstract
This paper investigates the incentive provision and selection properties of multi-stage contests with heterogeneous agents. We first derive analytical results for heterogeneous multi-stage contests and show that the allocation of player types in the first stage matters for both incentive provision and selection. In fact, a tournament designer faces a trade-off between those two goals. Further, we find that the incentive provision is always lower in heterogeneous than in homogenous multi-stage tournaments. Those implications are then tested in laboratory experiments. While we find substantial overprovision of effort compared to the theoretical benchmark in all treatments, cross treatment comparison provide strong support for all qualitative predictions in both the incentive provision and the selection dimension, in particular when the degree of heterogeneity between participants is high.