用經濟學實驗研究事前溝通如何改善求職歧視

Discrimination and Pre-game Communication: An Experimental Investigation

王鼎崴(Ding-Wei Wang)

指導教授 ： 王道一 ； 賴建宇

國立臺灣大學/社會科學院/經濟學系/碩士(2024年)

https://doi.org/10.6342/NTU202403397

摘要

我們用經濟學實驗研究求職者決定是否針對雇主投入受訓、雇主決定是否錄取求職者的應徵聘僱獵鹿賽局。我們在種子階段及基準階段根據 Dianat et al. (2022)的設計，誘發雇主對求職者的統計歧視，將求職者分為 GREEN（優勢族群）與 PURPLE（劣勢族群）。我們在隨後的政策階段評估以下三種反歧視政策：不讓雇主知悉求職者類別（「盲選」）、允許求職者主動揭露自身類別（「自我揭露」）或者透過付費傳達投入受訓之意願（「付費傳達」）。我們發現盲選政策能弭平族群歧視，但會損及優勢族群的權益，因此優勢族群會主動揭露自身類別來抵銷盲選的效果。付費傳達願意投入受訓的訊息能促進勞雇雙方合作達到「受訓/雇用」的均衡，因此廣為劣勢族群所使用。這些實驗結果點出盲選政策的潛在問題，而付費傳達則是可以考慮的替代方案。

關鍵字

刻板印象 ； 貼標籤 ； 汙名化 ； 協調賽局 ； 平權行動 ； 盲測 ； 職場訊息傳遞 ； 經濟學實驗

並列摘要

We conduct a Stag Hunt hiring experiment where workers decide on firm-specific investments and firms decide whether to hire the worker. In the Seed Stage and Baseline Stage, we follow Dianat et al. (2022) to induce statistical discrimination across workers, who can be GREEN (advantaged) or PURPLE (disadvantaged). Subsequently, in the Policy Stage, we evaluate the efficacy of three policy interventions: removing firm access to worker types (“Colorblind”), allowing workers to voluntarily disclose (“Type Disclosure”) or send a costly signal indicating their intention to invest (“Costly Signaling”). We find equalization occurs under colorblind policies at the expense of advantaged workers, but is undone by their voluntary type disclosure. Costly signal induces coordination on the investing-hiring equilibrium and is utilized mostly by disadvantaged workers. These findings highlight the potential drawbacks of colorblind policies and point to costly signaling as a less problematic alternative.

並列關鍵字

Stereotypes ； Labeling ； Stigmatization ； Coordination Games ； Affirmative Action ； Blind Audition ； Job Market Signaling ； Laboratory Experiments

參考文獻

American Bar Association, “Impact of Scotus Affirmative Action Ruling on Employers,” 2023.

Benndorf, Volker, Dorothea Kübler, and Hans-Theo Normann, “Privacy concerns, voluntary disclosure of information, and unraveling: An experiment,” European Economic Review, 2015, 75, 43–59.

Blume, Andreas, Peter H Kriss, and Roberto A Weber, “Pre-play communication with forgone costly messages: experimental evidence on forward induction,” Experimental Economics, 2017, 20, 368–395.

Chen, Daniel L, Martin Schonger, and Chris Wickens, “oTree—An open-source platform for laboratory, online, and field experiments,” Journal of Behavioral and Experimental Finance, 2016, 9, 88–97.

Dianat, Ahrash, Federico Echenique, and Leeat Yariv, “Statistical discrimination and affirmative action in the lab,” Games and Economic Behavior, 2022, 132, 41–58.