On Statistical Discrimination as a Failure of Social Learning: A Multiarmed Bandit Approach

References

• Abbasi-Yadkori Y, Pál D, Szepesvári C (2011) Improved algorithms for linear stochastic bandits. Taylor JS, Zemel RS, Bartlett PL, Pereira FCN, Weinberger KQ, eds. Proc. Twenty-fourth Conf. Neural Inform. Processing Systems (Curran Associates, New York), 2312–2320.Google Scholar
• Abe N, Long PM (1999) Associative reinforcement learning using linear probabilistic concepts. Bratko I, Dzeroski S, eds. Proc. Sixteenth Internat. Conf. Machine Learn. (Morgan Kaufmann, San Francisco), 3–11.Google Scholar
• Al-Ali MN (2004) How to get yourself on the door of a job: A cross-cultural contrastive study of Arabic and English job application letters. J. Multilingual Multicultural Development 25(1):1–23.Crossref, Google Scholar
• Arrow K (1973) The theory of discrimination. Ashenfelter O, Rees A, eds. Discrimination in Labor Markets (Princeton University Press, Princeton, NJ), 3–33.Google Scholar
• Auer P, Cesa-Bianchi N, Fischer P (2002) Finite-time analysis of the multi-armed bandit problem. Machine Learn. 47(2):235–256.Crossref, Google Scholar
• Banerjee AV (1992) A simple model of herd behavior. Quart. J. Econom. 107(3):797–817.Crossref, Google Scholar
• Bardhi A, Guo Y, Strulovici B (2020) Early-career discrimination: Spiraling or self-correcting? Working paper, Duke University, Durham, NC.Google Scholar
• Bastani H, Bayati M, Khosravi K (2021) Mostly exploration-free algorithms for contextual bandits. Management Sci. 67(3):1329–1349.Link, Google Scholar
• Bechavod Y, Ligett K, Roth A, Waggoner B, Wu SZ (2019) Equal opportunity in online classification with partial feedback. Wallach HM, Larochelle H, Beygelzimer A, d’Alché-Buc F, Fox EB, Garnett R, eds. Proc. Thirty-Third Conf. Neural Inform. Processing Systems (Curran Associates, New York), 8972–8982.Google Scholar
• Bikhchandani S, Hirshleifer D, Welch I (1992) A theory of fads, fashion, custom, and cultural change as informational cascades. J. Political Econom. 100(5):992–1026.Crossref, Google Scholar
• Bohren JA, Imas A, Rosenberg M (2019) The dynamics of discrimination: Theory and evidence. Amer. Econom. Rev. 109(10):3395–3436.Crossref, Google Scholar
• Bohren JA, Haggag K, Imas A, Pope DG (2023) Inaccurate statistical discrimination: An identification problem. Rev. Econom. Statist., 1–45.Crossref, Google Scholar
• Calders T, Verwer S (2010) Three naive Bayes approaches for discrimination-free classification. Data Min. Knowl. Discov. 21(2):277–292.Crossref, Google Scholar
• Che YK, Kim K, Zhong W (2019) Statistical discrimination in ratings-guided markets. Working paper, Columbia University, New York.Google Scholar
• Chen Y, Cuellar A, Luo H, Modi J, Nemlekar H, Nikolaidis S (2020) The fair contextual multi-armed bandit. Jonas P, David S, eds. Proc. Nineteenth Internat. Conf. Autonomous Agents Multiagent Systems (Journal of Machine Learning Research), 1810–1812.Google Scholar
• Chu W, Li L, Reyzin L, Schapire R (2011) Contextual bandits with linear payoff functions. Geoffrey G, David D, Miroslav D, eds. Proc. Fourteenth Internat. Conf. Artificial Intelligence Statist. (Journal of Machine Learning Research), 208–214.Google Scholar
• Coate S, Loury GC (1993) Will affirmative-action policies eliminate negative stereotypes? Amer. Econom. Rev. 83(5):1220–1240.Google Scholar
• Cornell B, Welch I (1996) Culture, information, and screening discrimination. J. Political Econom. 104(3):542–571.Crossref, Google Scholar
• Dianat A, Echenique F, Yariv L (2022) Statistical discrimination and affirmative action in the laboratory. Games Econom. Behav. 132:41–58.Crossref, Google Scholar
• Foster D, Vohra R (1992) An economic argument for affirmative action. Rationality Soc. 4(2):176–188.Crossref, Google Scholar
• Frazier P, Kempe D, Kleinberg J, Kleinberg R (2014) Incentivizing exploration. Babaioff M, Conitzer V, Easley DA, eds. Proc. Fifteenth ACM Conf. Econom. Comput. (Association for Computing Machinery, New York), 5–22.Google Scholar
• Hanna RN, Linden LL (2012) Discrimination in grading. Amer. Econom. J. Econom. Policy 4(4):146–168.Crossref, Google Scholar
• Hannák A, Wagner C, Garcia D, Mislove A, Strohmaier M, Wilson C (2017) Bias in online freelance marketplaces: Evidence from TaskRabbit and Fiverr. Lee CP, Poltrock SE, Barkhuus L, Borges M, Kellogg WA, eds. Proc. 2017 ACM Conf. Comput. Supported Cooperative Work Soc. Comput. (Association for Computing Machinery, New York), 1914–1933.Google Scholar
• Hardt M, Price E, Srebro N (2016) Equality of opportunity in supervised learning. Lee DD, Sugiyama M, von Luxburg U, Guyon I, Garnett R, eds. Proc. Twenty-ninth Conf. Neural Inform. Processing Systems (Curran Associates, New York), 3315–3323.Google Scholar
• Hu L, Chen Y (2018) A short-term intervention for long-term fairness in the labor market. Champin PA, Gandon G, Lalmas M, Ipeirotis PG, eds. Proc. 2018 World Wide Web Conf. (Association for Computing Machinery, New York), 1389–1398.Google Scholar
• Immorlica N, Mao J, Slivkins A, Wu ZS (2020) Incentivizing exploration with selective data disclosure. Biró P, Hartline JD, Ostrovsky M, Procaccia AD, eds. Proc. Twenty-First ACM Conf. Econom. Comput. (Association for Computing Machinery, New York), 647–648.Google Scholar
• Johari R, Kamble V, Krishnaswamy AK, Li H (2018) Exploration vs. exploitation in team formation. Christodoulou G, Harks T, eds. Proc. Fourteenth Conf. Web Internet Econom., vol. 11316 (Springer, Berlin, Heidelberg), 452.Google Scholar
• Joseph M, Kearns M, Morgenstern JH, Roth A (2016) Fairness in learning: Classic and contextual bandits. Lee DD, Sugiyama M, von Luxburg U, Guyon I, Garnett R, eds. Proc. Twenty-Ninth Conf. Neural Inform. Processing Systems (Curran Associates, New York), 325–333.Google Scholar
• Kannan S, Roth A, Ziani J (2019) Downstream effects of affirmative action. Boyd D, Morgenstern JH, eds. Proc. Second Conf. Fairness Accountability Transparency (Association for Computing Machinery, New York), 240–248.Google Scholar
• Kannan S, Morgenstern JH, Roth A, Waggoner B, Wu ZS (2018) A smoothed analysis of the greedy algorithm for the linear contextual bandit problem. Bengio S, Wallach HM, Larochelle H, Grauman K, Cesa-Bianchi N, Garnett R, eds. Proc. Thirty-Second Conf. Neural Inform. Processing Systems (Curran Associates, New York), 2227–2236.Google Scholar
• Kannan S, Kearns M, Morgenstern J, Pai M, Roth A, Vohra R, Wu ZS (2017) Fairness incentives for myopic agents. Proc. 2017 ACM Conf. Econom. Comput. (Association for Computing Machinery, New York), 369–386.Google Scholar
• Kennedy P (2008) A Guide to Econometrics, 6th ed. (Wiley-Blackwell, Hoboken, NJ), 192–202.Google Scholar
• Kleinberg JM, Mullainathan S, Raghavan M (2017) Inherent trade-offs in the fair determination of risk scores. Papadimitriou CH, ed. Proc. Eighth Conf. Innovations Theoret. Comput. Sci. (Schloss Dagstuhl – Leibniz-Zentrum für Informatik, Dagstuhl), 43:1–43:23.Google Scholar
• Kremer I, Mansour Y, Perry M (2014) Implementing the ‘wisdom of the crowd’. J. Political Econom. 122(5):988–1012.Crossref, Google Scholar
• Lai T, Robbins H (1985) Asymptotically efficient adaptive allocation rules. Adv. Appl. Math. 6(1):4–22.Crossref, Google Scholar
• Langford J, Zhang T (2008) The epoch-greedy algorithm for contextual multi-armed bandits. Proc. Twentieth Conf. Neural Inform. Processing Systems (Curran Associates, New York), 817–824.Google Scholar
• Li D, Raymond L, Bergman P (2020) Hiring as exploration. NBER Working Paper No. 27736, National Bureau of Economic Research, Cambridge, MA.Google Scholar
• MacNell L, Driscoll A, Hunt AN (2015) What’s in a name: Exposing gender bias in student ratings of teaching. Innovative Higher Ed. 40(4):291–303.Crossref, Google Scholar
• Mailath GJ, Samuelson L, Shaked A (2000) Endogenous inequality in integrated labor markets with two-sided search. Amer. Econom. Rev. 90(1):46–72.Crossref, Google Scholar
• Makhlouf K, Zhioua S, Palamidessi C (2021) Machine learning fairness notions: Bridging the gap with real-world applications. Inform. Processing Management 58(5):102642.Crossref, Google Scholar
• Mansour Y, Slivkins A, Syrgkanis V (2020) Bayesian incentive-compatible bandit exploration. Oper. Res. 68(4):1132–1161.Link, Google Scholar
• Mitchell KM, Martin J (2018) Gender bias in student evaluations. PS Political Sci. Politics 51(3):648–652.Crossref, Google Scholar
• Monachou FG, Ashlagi I (2019) Discrimination in online markets: Effects of social bias on learning from reviews and policy design. Wallach HM, Larochelle H, Beygelzimer A, d’Alché-Buc F, Fox EB, Garnett R, eds. Proc. Thirty-Second Conf. Neural Inform. Processing Systems (Curran Associates, New York), 2145–2155.Google Scholar
• Moro A, Norman P (2004) A general equilibrium model of statistical discrimination. J. Econom. Theory 114(1):1–30.Crossref, Google Scholar
• Neumark D (2018) Experimental research on labor market discrimination. J. Econom. Literature 56(3):799–866.Crossref, Google Scholar
• Owen AB, Varian H (2020) Optimizing the tie-breaker regression discontinuity design. Electronic J. Statist. 14(2):4004–4027.Crossref, Google Scholar
• Papanastasiou Y, Bimpikis K, Savva N (2018) Crowdsourcing exploration. Management Sci. 64(4):1727–1746.Link, Google Scholar
• Pedreschi D, Ruggieri S, Turini F (2008) Discrimination-aware data mining. Li Y, Liu B, Sarawagi S, eds. Proc. Fourteenth ACM SIGKDD Internat. Conf. Knowledge Discovery Data Mining (Association for Computing Machinery, New York), 560–568.Google Scholar
• Peña VH, Lai TL, Shao QM (2008) Self-Normalized Processes: Limit Theory and Statistical Applications (Springer Science & Business Media, Berlin).Google Scholar
• Phelps ES (1972) The statistical theory of racism and sexism. Amer. Econom. Rev. 62(4):659–661.Google Scholar
• Precht K (1998) A cross-cultural comparison of letters of recommendation. English Specific Purposes 17(3):241–265.Crossref, Google Scholar
• Raghavan M, Slivkins A, Vaughan JW, Wu ZS (2018) The externalities of exploration and how data diversity helps exploitation. Bubeck S, Perchet V, Rigollet P, eds. Proc. Machine Learn. Res., vol. 75 (PMLR), 1724–1738.Google Scholar
• Robbins H (1952) Some aspects of the sequential design of experiments. Bull. Amer. Math. Soc. (N.S.) 58(5):527–535.Crossref, Google Scholar
• Rusmevichientong P, Tsitsiklis JN (2010) Linearly parameterized bandits. Math. Oper. Res. 35(2):395–411.Link, Google Scholar
• Smith L, Sørensen P (2000) Pathological outcomes of observational learning. Econometrica 68(2):371–398.Crossref, Google Scholar
• Thompson WR (1933) On the likelihood that one unknown probability exceeds another in view of the evidence of two samples. Biometrika 25(3–4):285–294.Crossref, Google Scholar
• Trix F, Psenka C (2003) Exploring the color of glass: Letters of recommendation for female and male medical faculty. Discourse Soc. 14(2):191–220.Crossref, Google Scholar
• Xu L, Honda J, Sugiyama M (2018) A fully adaptive algorithm for pure exploration in linear bandits. Perez-Cruz, ed. Proc. Twenty-First Internat. Conf. Artificial Intelligence Statist. (Journal of Machine Learning Research), 843–851.Google Scholar