Technology is neither good nor bad; nor is it neutral (Kranzberg, 1986).
Abstract
With their intensive use of data to classify and price risk, insurers have often been confronted with data-related issues of fairness and discrimination. This paper provides a comparative review of discrimination issues raised by traditional statistics versus machine learning in the context of insurance. We first examine historical contestations of insurance classification, showing that it was organized along three types of bias: pure stereotypes, non-causal correlations, or causal effects that a society chooses to protect against, are thus the main sources of dispute. The lens of this typology then allows us to look anew at the potential biases in insurance pricing implied by big data and machine learning, showing that despite utopic claims, social stereotypes continue to plague data, thus threaten to unconsciously reproduce these discriminations in insurance. To counter these effects, algorithmic fairness attempts to define mathematical indicators of non-bias. We argue that this may prove insufficient, since as it assumes the existence of specific protected groups, which could only be made visible through public debate and contestation. These are less likely if the right to explanation is realized through personalized algorithms, which could reinforce the individualized perception of the social that blocks rather than encourages collective mobilization.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
This distinction is not always easy to make: see Charpentier et al. (2022) for a discussion in the case of natural disasters. Please note that we limit here the discussion to causal variables; Loi and Christen (2021) further discuss moral issues associated with choice when the variable is used as a surrogate.
By contrast, the 2018 European General Data Protection Regulation (GDPR) includes ethnicity among protected data, together with religious beliefs, sexual orientation, trade union involvement, medical status, criminal convictions and offences, biometric data and genetic information.
Heen (2009) argues that it is possible that in some southern US states, old life insurance policies from the Jim Crow period (i.e., using race as an underwriting parameter) were still in force in 2009.
In a more general context, Seele et al. (2021) define price personalization as charging customers differentially according to their willingness to pay. While such issues also exist in insurance (Lukacs et al., 2016), this paper is narrowly focused on the personalization of the risk premium, that is to say on the capacity to individually predict insurance costs.
Called “The Great AI Debate: Interpretability is necessary for machine learning”, opposing Rich Caruana and Patrice Simard (for) to Kilian Weinberger and Yann LeCun (against), https://youtu.be/93Xv8vJ2acI.
As Giovanola and Tiribelli (2022, p. 2) aptly note, the literature on algorithmic fairness limits the concept to the absence of bias. In their viewpoint, “it is questionable whether focusing exclusively on biases can encompass the complexity of the concept of fairness” (Giovanola & Tiribelli, 2022, p. 4). Similarly, focusing only on the absence of bias in insurance could lead to unfair situations where there are no bias but vulnerable people cannot afford insurance.
References
Ananny, M., & Crawford, K. (2018). Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability. New Media & Society, 20(3), 973–989. https://doi.org/10.1177/1461444816676645
Anderson, C. (2008). The end of theory: The data deluge makes the scientific method obsolete. Wired. https://www.wired.com/2008/06/pb-theory/
Austin, R. (1983). The insurance classification controversy. University of Pennsylvania Law Review, 131(3), 517–582. https://doi.org/10.2307/3311844
Avraham, R. (2017). Discrimination and insurance SSRN Scholarly Paper ID 3089946. Rochester, NY: Social Science Research Network. https://papers.ssrn.com/abstract=3089946
Baeder, L., Brinkmann, P., & Xu, E. (2021). Interpretable machine learning for insurance. Innovation and Technology. Society of Actuaries. https://www.soa.org/globalassets/assets/files/resources/research-report/2021/interpretable-machine-learning.pdf
Baker, T., & Simon, J. (2002). Embracing risk. In Embracing risk: The changing culture of insurance and responsibility (pp. 1–25). University of Chicago Press.
Barocas, S., & Selbst, A. D. (2016). Big data’s disparate impact essay. California Law Review, 104, 671–732.
Barry, L. (2020). Insurance, big data and changing conceptions of fairness. European Journal of Sociology / Archives Européennes De Sociologie, 61(2), 159–184. https://doi.org/10.1017/S0003975620000089
Barry, L. (2022). Epidemic and insurance: Two forms of solidarity. Theory, Culture & Society. https://doi.org/10.1177/02632764221087932
Barry, L., & Charpentier, A. (2020). Personalization as a promise: Can big data change the practice of insurance? Big Data & Society. https://doi.org/10.1177/2053951720935143
Barry, L., & Fisher, E. (2019). Digital audiences and the deconstruction of the collective. Subjectivity, 12(3), 210–227. https://doi.org/10.1057/s41286-019-00073-w
Bouk, D. (2015). How our days became numbered: Risk and the rise of the statistical individual. University Of Chicago Press.
boyd, d., & Crawford, K. (2012). Critical questions for big data. Information, Communication and Society, 15(5), 662–679. https://doi.org/10.1080/1369118X.2012.678878
Breiman, L. (2001). Statistical modeling: The two cultures (with comments and a rejoinder by the author). Statistical Science, 16(3), 199–231. https://doi.org/10.1214/ss/1009213726
Caliskan, A., Bryson, J. J., & Narayanan, A. (2017). Semantics derived automatically from language corpora contain human-like biases. Science, 356(6334), 183–186. https://doi.org/10.1126/science.aal4230
Caruana, R., Lou, Y., Gehrke, J., Koch, P., Sturm, M., & Elhadad, N. (2015). Intelligible models for HealthCare: Predicting pneumonia risk and hospital 30-day readmission. In Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining, 1721–30. KDD ’15. New York, NY, USA: Association for Computing Machinery. https://doi.org/10.1145/2783258.2788613
CAS Race and Insurance Research Task Force. (2021). Approaches to address racial bias. CAS Research Paper Series on Race and Insurance Pricing.
Cevolini, A., & Esposito, E. (2020). From pool to profile: Social consequences of algorithmic prediction in insurance. Big Data & Society, 7(2), 1–11. https://doi.org/10.1177/2053951720939228
Charpentier, A. (2022). Insurance: Discrimination, biases and fairness. Opinions and debates. Institut Louis Bachelier. https://www.institutlouisbachelier.org/en/insurance-discrimination-biases-fairness/.
Charpentier, A., Barry, L., & Gallic, E. (2020). Quel avenir pour les probabilités prédictives en assurance ? Annales Des Mines - Realites Industrielles, 2020(1), 74–77.
Charpentier, A., Barry, L., & James, M. R. (2022). Insurance against natural catastrophes: Balancing actuarial fairness and social solidarity. The Geneva Papers on Risk and Insurance - Issues and Practice, 47(1), 50–78. https://doi.org/10.1057/s41288-021-00233-7
Charpentier, A., Denuit, M. M., & Elie, R. (2015). Segmentation et Mutualisation, Les Deux Faces d’une Même Pièce. Risques, 103, 19–23.
Charpentier, A., Hu, F., & Ratz, P. (2023). Mitigating discrimination in insurance with Wasserstein Barycenters. arXiv. http://arxiv.org/abs/2306.12912
De Wit, G. W., & Van Eeghen, J. (1984). Rate making and society’s sense of fairness. ASTIN Bulletin, 14(2), 151–164.
De Pril, N., & Dhaene, J. (1996). Segmentering in Verzekeringen. KUL. Departement toegepaste economische wetenschappen.
Desrosières, A. (2008). L’argument statistique. I, Pour une sociologie historique de la quantification. Presses de l’école des Mines.
Dworkin, R. (1981). What is equality? Part 2: Equality of resources. Philosophy & Public Affairs, 10(4), 283–345.
Ewald, F. (1986). L’Etat Providence. Grasset.
Ewald, F. (2011). Omnes et Singulatim. After risk. Carceral Notebooks, 7, 77–107.
Fisher, R. A. (1971). The design of experiments. Macmillan Pub Co.
Foucault, M. (1995). Discipline & punish: The birth of the prison (A. Sheridan, Trans.). Vintage Books.
François, P., & Voldoire, T. (2022). The revolution that did not happen. Telematics and car insurance in the 2010s. Big Data & Society, 9(2). https://doi.org/10.1177/20539517221142033
Frees, E. W., & Huang, F. (2021). The discriminating (pricing) actuary. North American Actuarial Journal. https://doi.org/10.1080/10920277.2021.1951296
Frezal, S., & Barry, L. (2020). Fairness in uncertainty: Some limits and misinterpretations of actuarial fairness. Journal of Business Ethics, 167(1), 127–136. https://doi.org/10.1007/s10551-019-04171-2
Giovanola, B., & Tiribelli, S. (2022). Weapons of moral construction? On the value of fairness in algorithmic decision-making. Ethics and Information Technology, 24(1), 3. https://doi.org/10.1007/s10676-022-09622-5
Glenn, B. J. (2000). The shifting rhetoric of insurance denial. Law & Society Review, 34(3), 779–808. https://doi.org/10.2307/3115143
Glenn, B. J. (2003). Postmodernism: The basis of insurance. Risk Management & Insurance Review, 6(2), 131–143. https://doi.org/10.1046/J.1098-1616.2003.028.x
Heen, M. (2009). Ending Jim Crow life insurance rates. Northwestern Journal of Law & Social Policy, 4(2), 360–399.
Heller, D. (2015). High price of mandatory auto insurance in predominantly African American Communities · Consumer Federation of America. Consumer Federation of America. https://consumerfed.org/reports/high-price-of-mandatory-auto-insurance-in-predominantly-african-american-communities/
Horan, C. D. (2021). Insurance era: Risk, governance, and the privatization of security in Postwar America (1st ed.). Chicago: University of Chicago Press.
Kita-Wojciechowska, K., & Kidziński, L. (2019). Google street view image predicts car accident risk. Central European Economic Journal, 6(53), 152–163.
Kiviat, B. (2019). The moral limits of predictive practices: The case of credit-based insurance scores. American Sociological Review, 84(6), 1134–1158. https://doi.org/10.1177/0003122419884917
Kleinberg, J., Mullainathan, S., & Raghavan, M. (2016). Inherent trade-offs in the fair determination of risk scores, September. https://arxiv.org/abs/1609.05807v2
Kranzberg, M. (1986). Technology and history: ‘Kranzberg’s Laws.’ Technology and Culture, 27(3), 544–560. https://doi.org/10.2307/3105385
Krippner, G. R., & Hirschman, D. (2022). The person of the category: The pricing of risk and the politics of classification in insurance and credit. Theory and Society. https://doi.org/10.1007/s11186-022-09500-5
Kusner, M. J., & Loftus, J. R. (2020). The long road to fairer algorithms. Nature, 578(7793), 34–36. https://doi.org/10.1038/d41586-020-00274-3
Larson, J., Angwin, J., Kirchner, L., & Mattu, S. (2017). How we examined racial discrimination in auto insurance prices. ProPublica. https://www.propublica.org/article/minority-neighborhoods-higher-car-insurance-premiums-methodology?token=oXaDaCvsdX3ZY7-YJd8F3L-6fSTJ6BUj
LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444. https://doi.org/10.1038/nature14539
Lehtonen, T.-K., & Liukko, J. (2011). The forms and limits of insurance solidarity. Journal of Business Ethics, 103(1), 33–44. https://doi.org/10.1007/s10551-012-1221-x
Leicht-Deobald, U., Busch, T., Schank, C., Weibel, A., Schafheitle, S., Wildhaber, I., & Kasper, G. (2019). The challenges of algorithm-based HR decision-making for personal integrity. Journal of Business Ethics, 160(2), 377–392. https://doi.org/10.1007/s10551-019-04204-w
Leigh, A. (2018). Randomistas: How radical researchers are changing our world. Yale University Press.
Lemonade. (2021). Lemonade’s claims automation. Lemonade Blog (blog). https://www.lemonade.com/blog/lemonades-claim-automation/
Lindholm, M., Richman, R., Tsanakas, A., & Wüthrich, M. (2022a). Discrimination free insurance pricing. ASTIN Bulletin: THe Journal of the IAA, 52(1), 55–89. https://doi.org/10.1017/asb.2021.23
Lindholm, M., Richman, R., Tsanakas, A., & Wüthrich, M. (2022b). A discussion of discrimination and fairness in insurance pricing. SSRN scholarly paper. Rochester, NY. https://doi.org/10.2139/ssrn.4207310
Lippert-Rasmussen, K. (2007). Nothing personal: On statistical discrimination. Journal of Political Philosophy, 15(4), 385–403. https://doi.org/10.1111/j.1467-9760.2007.00285.x
Loi, M., & Christen, M. (2021). Choosing how to discriminate: Navigating ethical trade-offs in fair algorithmic design for the insurance sector. Philosophy & Technology, 34(4), 967–992. https://doi.org/10.1007/s13347-021-00444-9
Lukacs, P., Neubecker, L., & Rowan, P. (2016). Price discrimination and cross-subsidy in financial services. Financial Conduct Authority Occasional Papers in Financial Regulation, Occasional Papers, No. 22.
Lury, C., & Day, S. (2019). Algorithmic personalization as a mode of individuation. Theory, Culture & Society, 36(2), 17–37. https://doi.org/10.1177/0263276418818888
Lyons, H., Velloso, E., & Miller, T. (2021). Conceptualising contestability: Perspectives on contesting algorithmic decisions. Proceedings of the ACM on Human-Computer Interaction, 5(CSCW1), 1–25. https://doi.org/10.1145/3449180
Martin, K. (2019). Ethical implications and accountability of algorithms. Journal of Business Ethics, 160(4), 835–850. https://doi.org/10.1007/s10551-018-3921-3
Mayer-Schönberger, V., & Cukier, K. (2014). Big data: A revolution that will transform how we live, work, and think (Reprint). Eamon Dolan/Mariner Books.
Mazzanti, S. (2021). SHAP explained the way I wish someone explained it to me. Towards Datascience (blog). April 21, 2021. https://towardsdatascience.com/shap-explained-the-way-i-wish-someone-explained-it-to-me-ab81cc69ef30
McFall, L., Meyers, G., & Van Hoyweghen, I. (2020). Editorial: The personalisation of insurance: Data, behaviour and innovation. Big Data & Society, 7(2), 2053951720973707. https://doi.org/10.1177/2053951720973707
McFall, L., & Moor, L. (2018). Who, or what, is insurtech personalizing?: Persons, prices and the historical classifications of risk. Distinktion: Journal of Social Theory, 19(2), 193–213. https://doi.org/10.1080/1600910X.2018.1503609
Metz, J. (2020). Sen. Booker’s PAID act looks to eliminate discriminatory non-driving factors in auto insurance pricing. Forbes Advisor. October 5, 2020. https://www.forbes.com/advisor/car-insurance/paid-act/
Meyers, G. (2018). Behaviour-based personalisation in health insurance: A sociology of a not-yet market. PhD Thesis, KU Leuven. https://limo.libis.be/primo-explore/fulldisplay?docid=LIRIAS2087689&context=L&vid=Lirias&search_scope=Lirias&tab=default_tab&lang=en_US&fromSitemap=1
Meyers, G., & Van Hoyweghen, I. (2018). Enacting actuarial fairness in insurance: From fair discrimination to behaviour-based fairness. Science as Culture, 27(4), 413–438. https://doi.org/10.1080/09505431.2017.1398223
Moor, L., & Lury, C. (2018). Price and the person: Markets, discrimination, and personhood. Journal of Cultural Economy, 11(6), 501–513. https://doi.org/10.1080/17530350.2018.1481878
Moriarty, J. (2021). Why online personalized pricing is unfair. Ethics and Information Technology, 23(3), 495–503. https://doi.org/10.1007/s10676-021-09592-0
Morrison, S. (2021). A disturbing, viral twitter thread reveals how AI-powered insurance can go wrong. Vox. May 27, 2021. https://www.vox.com/recode/22455140/lemonade-insurance-ai-twitter
Mosley, R., & Wenman, R. (2021). Methods for quantifying discriminatory effects on protected classes in insurance. CAS research paper series on race and insurance pricing, 26.
Napoletani, D., Panza, M., & Struppa, D. C. (2011). Agnostic science. Towards a philosophy of data analysis. Foundations of Science, 16(1), 1–20. https://doi.org/10.1007/s10699-010-9186-7
O’Neil, C. (2016). Weapons of math destruction: How big data increases inequality and threatens democracy (1st ed.). Crown.
Pasquale, F. (2016). The Black Box Society: The secret algorithms that control money and information (Reprint). Harvard University Press.
Pearl, J., & Mackenzie, D. (2018). The book of why: The new science of cause and effect (1st ed.). Basic Books.
Pessach, D., & Shmueli, E. (2020). Algorithmic fairness, https://arxiv.org/abs/2001.09784v1
Pethig, F., & Kroenung, J. (2022). Biased humans, (un)biased algorithms? Journal of Business Ethics, 83 637-652. https://doi.org/10.1007/s10551-022-05071-8
Prince, A. E. R., & Schwarcz, D. (2019). Proxy discrimination in the age of artificial intelligence and big data. Iowa Law Review, 105, 1257–1318.
Rebert, L., & Van Hoyweghen, I. (2015). The right to underwrite gender. The goods & services directive and the politics of insurance pricing. Tijdschrift Voor Genderstudies, 18(4), 413–431.
Regalbuto, J. (2019). Insurance Circular Letter No. 1 (2019): Use of external consumer data and information sources in underwriting for life insurance. Department of Financial Services. https://www.dfs.ny.gov/industry_guidance/circular_letters/cl2019_01
Rice, P. (2019). DFS Circular Letter No. 1 requires life insurers to justify use of external data sources and models in life underwriting | Insights | DLA Piper Global Law Firm. DLA Piper. https://www.dlapiper.com/en/france/insights/publications/2019/02/dfs-circular-letter/
Rosenbaum, P. (2017). Observation and experiment: An introduction to causal inference. Harvard University Press. https://doi.org/10.4159/9780674982697
Schauer, F. (2003). Profiles, probabilities, and stereotypes. Harvard University Press. https://doi.org/10.2307/j.ctvjz82xm
Schreiber, D. (2018). AI Eats Insurance. Lemonade Blog (blog). https://www.lemonade.com/blog/ai-eats-insurance/
Seele, P., Dierksmeier, C., Hofstetter, R., & Schultz, M. D. (2021). Mapping the ethicality of algorithmic pricing: A review of dynamic and personalized pricing. Journal of Business Ethics, 170(4), 697–719. https://doi.org/10.1007/s10551-019-04371-w
Shikhare, S. (2021). AI enabled next generation LTC and life insurance underwiring using facial score model. In Insurance data science conference 2021, 19. London. https://insurancedatascience.org/downloads/London2021/Session_4b/Shrinivas_Shikhare.pdf
Simon, J. (1988). The ideological effects of actuarial practices. Law Social Review, 22, 771–800.
Swedloff, R. (2014). Risk classification big data (r)evolution. Connecticut Insurance Law Journal, 21(1), 339–373.
Tanninen, M. (2020). Contested technology: Social scientific perspectives of behaviour-based insurance. Big Data & Society, 7(2), 2053951720942536. https://doi.org/10.1177/2053951720942536
Thiery, Y., & Van Schoubroeck, C. (2006). Fairness and equality in insurance classification. The Geneva Papers on Risk and Insurance - Issues and Practice, 31(2), 190–211. https://doi.org/10.1057/palgrave.gpp.2510078
Tselentis, D. I., Yannis, G., & Vlahogianni, E. I. (2017). Innovative motor insurance schemes: A review of current practices and emerging challenges. Accident Analysis & Prevention, 98(January), 139–148. https://doi.org/10.1016/j.aap.2016.10.006
Walters, M. A. (1981). Risk classification standards. Proceedings of the Casualty Actuarial Society, 68, 1–23.
Wiggins, B. A. (2013). Managing risk, managing race: racialized actuarial science in the United States, 1881–1948. Minnesota. http://conservancy.umn.edu/handle/11299/159587
Williams, B. A., Brooks, C. F., & Shmargad, Y. (2018). How algorithms discriminate based on data they lack: Challenges, solutions, and policy implications. Journal of Information Policy, 8, 78–115. https://doi.org/10.5325/jinfopoli.8.2018.0078
Wolff, M. J. (2006). The myth of the actuary: Life insurance and Frederick L. Hoffman’s race traits and tendencies of the American Negro. Public Health Reports, 121(1), 84–91.
Works, R. (1977). Whatever’s FAIR—Adequacy, equity, and the underwriting prerogative in property insurance markets. Nebraska Law Review, 56(3), 445–464.
Xin, X., & Huang, F. (2022). Anti-discrimination insurance pricing: Regulations, fairness criteria, and models. SSRN scholarly paper. Rochester, NY. https://doi.org/10.2139/ssrn.3850420
Funding
This study was supported by the Chaire PARI – project ‘Evaluation des risques et technologies du big data: Outils et conséquences’, Fondation Institut Europlace de Finance.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The authors have no competing interests to declare that are relevant to the content of this article.
Data availability
Not applicable.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Barry, L., Charpentier, A. Melting contestation: insurance fairness and machine learning. Ethics Inf Technol 25, 49 (2023). https://doi.org/10.1007/s10676-023-09720-y
Accepted:
Published:
DOI: https://doi.org/10.1007/s10676-023-09720-y