Why Informationally Diverse Teams Need Not Form, Even When Efficient

37 Pages Posted: 31 Jan 2018 Last revised: 28 Oct 2022

See all articles by Ashwin Kambhampati

Ashwin Kambhampati

United States Naval Academy

Carlos Segura-Rodriguez

Central Bank of Costa Rica

Peng Shao

University of Pennsylvania

Date Written: October 20, 2022

Abstract

In recent decades, research organizations have brought the “market inside the firm” by allowing workers to sort themselves into teams. How do teams form absent a central authority? We introduce a model of team formation in which workers first match and then produce correlated signals about an unknown state. While it is efficient to maximize the number of informationally diverse teams, such teams need not form in equilibrium. Our analysis identifies the two sources of matching inefficiency: (i) workers may form diverse teams that are beneficial to its members, but force excluded workers to form homogeneous teams, and (ii) even when a diverse team is efficient, a worker may prefer to join a homogeneous team if she can exert less effort than her teammate. We completely characterize each inefficiency, discuss contracts that restore efficiency, and relate these contracts to observed management practice.

Keywords: Matching, Teams, Information Acquisition, Correlation

JEL Classification: C78, L23, D83

Suggested Citation

Kambhampati, Ashwin and Segura-Rodriguez, Carlos and Shao, Peng, Why Informationally Diverse Teams Need Not Form, Even When Efficient (October 20, 2022). Available at SSRN: https://ssrn.com/abstract=3113594 or http://dx.doi.org/10.2139/ssrn.3113594

Ashwin Kambhampati

United States Naval Academy ( email )

121 Blake Road
Annapolis, MD 21402
United States

Carlos Segura-Rodriguez (Contact Author)

Central Bank of Costa Rica ( email )

Apartado Postal 10058
1000 San Jose
United States

Peng Shao

University of Pennsylvania ( email )

Philadelphia, PA 19104
United States

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

Paper statistics

Downloads
151
Abstract Views
1,242
Rank
353,993
PlumX Metrics