Joint vs. Separate Crowdsourcing Contests

• 成果类型: Article
• 署名作者: Hu, Ming; Wang, Lu
• 署名单位: University of Toronto; Shanghai University of Finance & Economics
• 刊物名称: MANAGEMENT SCIENCE
• ISSN/ISSBN: 0025-1909
• DOI: 10.1287/mnsc.2020.3683
• 发表日期: 2021
• 页码: 2711-2728
• 关键词: Game theory; crowdsourcing contest; Tournament design

摘要

In a crowdsourcing contest, innovation is outsourced by a firm to an open crowd that competes in generating innovative solutions. Given that the projects typically consist of multiple attributes, how should the firm optimally design a crowdsourcing contest for such a project? We consider two alternative mechanisms. One is a joint contest, where the best solution is chosen from the joint solutions across attributes submitted by all contestants. The other ismultiple separate parallel subcontests, with each dedicated to one attribute of the project. It is intuitive that the separate contest has the advantage of potentially creating a cooperative final solution contributed by different contestants. However, somewhat surprisingly, we show that the separate contest may reduce the incentive for the crowd to exert effort, resulting in the joint contest becoming the optimal scheme. The comparison of the expected best performances in the two contests depends on the project's characteristics. For example, if contestants' performances have a sufficiently high (respectively, low) level of randomness, the separate (respectively, joint) contest is optimal. If thenumber of contestants is large (respectively, small) enough, the separate (respectively, joint) contest is optimal. Moreover, we find that when the prize is endogenized, the optimal amount of the prize in the joint contest is no less than that in the separate contest. Finally, we extend the model to account for contestants with heterogeneous types.