Entry and disclosure in group contests

Luke Boosey, Philip Brookins, Dmitry Ryvkin · 2025

Read Entry and disclosure in group contests by Luke Boosey, Philip Brookins, Dmitry Ryvkin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-20092.txt.

Introduction

In group contests, a team's score is often its best member's effort. Luke Boosey, Philip Brookins, and Dmitry Ryvkin ask what organizers should reveal, and when. Players first decide whether to enter, holding private values. Organizers can hide entry information, share it inside each group, or broadcast it to all. The setting is an all-pay contest where the top group wins for sure.

For solo contests, disclosure always cuts total effort. Groups flip the result. Sharing information inside the group clearly raises total investment. Broadcasting it to everyone has an unclear effect. The authors show that team structure changes the value of transparency itself.

Catalog record

CREATORLuke Boosey, Philip Brookins, Dmitry Ryvkin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE101 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-20092

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Entry and disclosure in group contests Thanks: We thank three anonymous reviewers for their insightful comments, as well as participants of the 5th annual conference on Contests: Theory and Evidence and the 95th annual meeting of the Southern Economic Association . All the remaining errors are ours. Luke Boosey, Philip Brookins, Dmitry Ryvkin Abstract We study information disclosure policies for contests among groups. Each player decides whether or not to participate in competition as a member of their group. We consider a generalized all-pay auction setting, in which within-group aggregation of effort is best-shot and the group with the highest performance wins the contest. Players’ values for winning are private information at the entry stage, but may be disclosed at the competition stage. We compare three disclosure policies: (i) no disclosure, when the number of entrants remains unknown and their values private; (ii) within-group disclosure, when this information is disclosed within each group but not across groups; and (iii) full disclosure, when the information about entrants is disclosed across groups. For contests between individuals, information disclosure always reduces expected aggregate investment. However, this is no longer true in group contests: Within-group disclosure unambiguously raises aggregate investment, while the effect of full disclosure is ambiguous. Keywords : group contest, best shot, endogenous entry, information disclosure JEL classification codes : C72, D82 1 Introduction Across a wide range of economic, political, and social environments, competition takes place between groups of individuals who align themselves in pursuit of some common goals. For example, many interest groups engaged in lobbying activities consist of diverse collections of individuals with broadly aligned objectives. Within organizations, managers may solicit project submissions from multiple divisions or teams of employees and reward the team with the best proposal. Crowdsourcing R&D platforms such as the XPRIZE Foundation invite groups to compete for prizes solving complex innovation problems. Such competitive settings can be broadly categorized as group contests . In each group, individuals invest effort or other resources to increase their group’s performance, which in turn improves the group’s chances of winning a valuable prize. In this paper, we study competition between groups in which the decision to enter competition as a member of a particular group is endogenous. In our setting, there are a fixed number of competing groups, each with a pool of potential participants. In an initial entry stage, each player decides whether or not to participate in competition as a member of their group. Participants must forgo some outside option or, equivalently, face a cost of entry. Then, in a second stage, participants make investment or effort decisions that determine the group’s performance level. We analyze and compare three disclosure policies that dictate the information available to participants at the time they make their investments. We consider activities in which group performance is determined solely by the best-shot , i.e., the maximum investment chosen by a group member ( Baik and Shogren, 1998 ; Chowdhury, Lee and Sheremeta, 2013 ) . 1 1 1 Other popular approaches are to model the group performance level using perfect substitutes aggregation technology ( Katz, Nitzan and Rosenberg, 1990 ; Baik, 1993 ; Baik, 2008 ; Baik, Kim and Na, 2001 ; Münster, 2009 ) or weak-link (perfect complements) technology ( Lee, 2012 ) . Kolmar and Rommeswinkel (2013) and Brookins, Lightle and Ryvkin (2015) consider varying degrees of complementarity with a CES technology. Such a setting may arise, for example, in lobbying environments where the official who is lobbied considers only the individual agents on each side of an issue who made the most compelling case; or within an organization where a team of employees pitches only the most promising project idea to the management. One can think also of a market setting where a client, such as a real-estate developer, solicits projects from multiple firms, and each firm conducts an internal selection process and presents its best proposal. We model group competition as a generalized all-pay auction in which the group with the highest performance (the highest best-shot) wins the contest with certainty. 2 2 2 This setting corresponds to a perfectly discriminating contest success function (CSF). Alternative environments, in which the contest is imperfectly discriminating, include those with a lottery CSF of Tullock (1980) and its generalizations. In these settings, the group with the highest performance has a higher probability of winning, but does not win with certainty. Moreover, we assume that the prize associated with winning is a group-specific public good. Thus, the value of winning for each member of the winning group is equal to her individual private value, regardless of the size of the group. 3 3 3 Note that potential members of the winning group who do not participate do not receive any benefit from the prize being awarded to the group they could have joined. This is the case, for example, for researchers deciding whether or not to join their colleagues on a grant application, politicians joining various caucuses or factions, or (potential) plaintiffs in group litigation. Together, these features of the environment generate incentives for individuals within groups to free-ride on the investments made by their fellow group members. These free-riding incentives underscore one of the key differences between contests among groups and contests among individuals. As such, the environment we consider provides a rich and previously unexplored interplay between endogenous entry, free riding and information disclosure in contests. The game proceeds in two stages. At the initial entry stage, players’ values for the prize are private information, although it is common knowledge that they are drawn independently from the same commonly known distribution. Depending on the disclosure policy, information about the number of participating group members and their values may become known to participants at the beginning of the second stage, prior to their investment decisions. 4 4 4 We assume this disclosure of information is exogenous. It may be the effect of explicit design decisions, for example, by a contest sponsor; or it may be a result of naturally occurring dissemination of information due to environmental factors, such as the spatial or network structure of agents’ interactions. In the second stage, participants simultaneously choose an irreversible, costly investment, and the outcome of the contest is determined. We first consider, as a benchmark, the case of contests among individuals and show that expected aggregate investment is always lower when information about others’ types is disclosed to those who enter. In this case, equilibrium entry is independent of disclosure. However, equilibrium aggregate investment corresponds to the (truncated) expectation of the second highest valuation without disclosure; and to the sum of the expected bids by the two highest valuation entrants in a complete information all-pay auction, with disclosure. The latter entails an efficiency loss because the lower valuation player can win, and a loss of revenue because the lower valuation player can bid zero, with a positive probability. There is also no scope for free riding or a coordination problem disclosure may alleviate; thus, nondisclosure leads to a larger investment. We then turn to contests among groups and consider three different disclosure rules. In the no disclosure (ND) setting, all entry decisions and valuations remain private information throughout the investment stage. As a

Read the complete text