Fairness of Ratemaking for Catastrophe Insurance: : Lessons from Machine Learning
Authors: 
Nan Zhang, Heng XuAuthors Info & Claims
Pages 469 - 488
Abstract
Catastrophe insurance is an important element of disaster management. Yet the historical presence of inequalities in insurance, from redlining to pricing disparity, has had a devastating impact on minority communities. Whereas the fairness of insurance ratemaking is studied in general, we identify a unique challenge for catastrophe insurance that sets it apart from other lines of insurance. Drawing upon the recent advances in machine learning for fair data valuation, we reveal striking connections between the two seemingly unrelated problems and lean on insights from machine learning to mathematically and empirically study the fairness of ratemaking methods for catastrophe insurance. Our results indicate the potential existence of disparate impact against minorities across existing methods and point to a unique mathematical solution that can satisfy a few commonly assumed properties of fair ratemaking for catastrophe insurance.
History: This paper has been accepted for the Information Systems Research Special Section on Unleashing the Power of Information Technology for Strategic Management of Disasters. Ahmed Abbasi, Robin Dillon-Merrill, H. Raghav Rao, and Olivia Sheng, Senior Editors; Guodong (Gordon) Gao, Associate Editor.
Funding: This work was supported in part by the National Science Foundation Division of Information and Intelligent Systems [Grant 2040807] and Amazon Science.
Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2022.1195.
References
[1]
Abbasi A, Li J, Clifford G, Taylor H (2018) Make “fairness by design” part of machine learning. Harvard Bus. Rev. https://hbr.org/2018/08/make-fairness-by-design-part-of-machine-learning.
[2]
Acerbi C, Tasche D (2002) Expected shortfall: A natural coherent alternative to value at risk. Econom. Notes 31(2):379–388.
[3]
Adjerid I, Peer E, Acquisti A (2018) Beyond the privacy paradox: Objective vs. relative risk in privacy decision making. Management Inform. Systems Quart. 42(2):465–488.
[4]
Akter S, Mallick B (2013) The poverty-vulnerability-resilience nexus: Evidence from Bangladesh. Ecological Econom. 96(1):114–124.
[5]
Alonso-Meijide JM, Álvarez-Mozos M, Fiestras-Janeiro MG, Jiménez-Losada A (2021) Marginality and convexity in partition function form games. Math. Methods Oper. Res. 94(1):99–121.
[6]
American Academy of Actuaries (2018) Uses of Catastrophe Model Output (American Academy of Actuaries, Washington, DC).
[7]
Angst CM, Block ES, D’arcy J, Kelley K (2017) When do it security investments matter? Accounting for the influence of institutional factors in the context of healthcare data breaches. Management Inform. Systems Quart. 41(3):893–916.
[8]
Arrow KJ (1978) Uncertainty and the welfare economics of medical care. Uncertainty in Economics (Academic Press, Cambridge, MA).
[9]
Artzner P, Delbaen F, Eber JM, Heath D (1999) Coherent measures of risk. Math. Finance 9(3):203–228.
[10]
Bachrach Y, Markakis E, Resnick E, Procaccia AD, Rosenschein JS, Saberi A (2010) Approximating power indices: Theoretical and empirical analysis. Autonomous Agents Multi-Agent Systems 20(2):105–122.
[11]
Baker T (2002) Risk, insurance, and the social construction of responsibility. Embracing Risk: The Changing Culture of Insurance and Responsibility, (University of Chicago Press, Chicago) 33–51.
[12]
Baker T (2011) Health insurance, risk, and responsibility after the Patient Protection and Affordable Care Act. Univ. Pennsylvania Law Rev. 159(6):1577–1622.
[13]
Barrett M, Walsham G (1999) Electronic trading and work transformation in the London insurance market. Inform. Systems Res. 10(1):1–22.
[14]
Bassellier G, Benbasat I, Reich BH (2003) The influence of business managers’ IT competence on championing it. Inform. Systems Res. 14(4):317–336.
[15]
Berman R (2018) Beyond the last touch: Attribution in online advertising. Marketing Sci. 37(5):771–792.
[16]
Bharosa N, Lee J, Janssen M (2010) Challenges and obstacles in sharing and coordinating information during multi-agency disaster response: Propositions from field exercises. Inform. Systems Frontiers 12(1):49–65.
[17]
Bourbaki N (2008) Lie Groups and Lie Algebras (Springer Science & Business Media, New York).
[18]
Bühlmann H (1970) Mathematical Methods in Risk Theory (Springer Science & Business Media, New York).
[19]
Cahn AL (2008) Our rights are not cast in stone: Post-Katrina environmental red-lining and the need for a broad-based human right lawyering movement. Univ. Pennsylvania J. Law Soc. Change 12(1):37–71.
[20]
Castro J, Gómez D, Tejada J (2009) Polynomial calculation of the Shapley value based on sampling. Comput. Oper. Res. 36(5):1726–1730.
[21]
Charpentier A, Barry L, James MR (2022) Insurance against natural catastrophes: Balancing actuarial fairness and social solidarity. Geneva Papers Risk Insurance-Issues Practice 47(1):50–78.
[22]
Colten CE (2006) Vulnerability and place: Flat land and uneven risk in New Orleans. Amer. Anthropologist 108(4):731–734.
[23]
Copeland TE, Weston JF, Shastri K (2005) Financial Theory and Corporate Policy, vol. 4 (Pearson, New York).
[24]
Cormen TH, Leiserson CE, Rivest RL, Stein C (2009) Introduction to Algorithms (MIT Press, Cambridge, MA).
[25]
Daniels N (1990) Insurability and the HIV epidemic: Ethical issues in underwriting. Milbank Quart. 68(4):497–525.
[26]
Deng X, Papadimitriou CH (1994) On the complexity of cooperative solution concepts. Math. Oper. Res. 19(2):257–266.
[27]
Dietz S, Niehörster F (2021) Pricing ambiguity in catastrophe risk insurance. Geneva Risk Insurance Rev. 46(2):112–132.
[28]
Dietz S, Walker O (2019) Ambiguity and insurance: Capital requirements and premiums. J. Risk Insurance 86(1):213–235.
[29]
Dong W, Shah H, Wong F (1996) A rational approach to pricing of catastrophe insurance. J. Risk Uncertainty 12(2):201–218.
[30]
Dwork C (2008) Differential privacy: A survey of results. Proc. Internat. Conf. Theory Appl. Models Comput., 1–19.
[31]
Embrechts P, Klüppelberg C, Mikosch T (2013) Modelling Extremal Events (Springer, Berlin).
[32]
Everett RS, Wojtkiewicz RA (2002) Difference, disparity, and race/ethnic bias in federal sentencing. J. Quant. Criminology 18(2):189–211.
[33]
Ewald F (2014) L’Etat Providence (Grasset, Paris).
[34]
Fatima SS, Wooldridge M, Jennings NR (2008) A linear approximation method for the Shapley value. Artificial Intelligence 172(14):1673–1699.
[35]
Fothergill A (2006) In the Wake of the Storm: Environment, Disaster, and Race After Katrina (SAGE, New York).
[36]
Frezal S, Barry L (2020) Fairness in uncertainty: Some limits and misinterpretations of actuarial fairness. J. Bus. Ethics 167(1):127–136.
[37]
Fu L, Khury C (2010) Optimal layers for catastrophe reinsurance. Variance 4(2):191–208.
[38]
Ghorbani A, Zou J (2019) Data Shapley: Equitable valuation of data for machine learning. Proc. Machine Learn. Res. 97(1):2242–2251.
[39]
Goode S, Hoehle H, Venkatesh V, Brown SA (2017) User compensation as a data breach recovery action: An investigation of the Sony Playstation network breach. Management Inform. Systems Quart. 41(3):703–727.
[40]
Grossi P (2005) Catastrophe Modeling: A New Approach to Managing Risk (Springer, New York).
[41]
Gul F (1989) Bargaining foundations of Shapley value. Econometrica 57(1):81–95.
[42]
Ha J, Kose MA, Ohnsorge F (2021) One-stop source: A global database of inflation. Policy Research Working Paper 9737, World Bank, Washington, DC.
[43]
Heller D, Advocate C (2015) High price of mandatory auto insurance in predominantly African American communities. Technical Report 2015-11-18, Consumer Federation of America, Washington, DC.
[44]
Herepath A (2014) In the loop: A realist approach to structure and agency in the practice of strategy. Organ. Stud. 35(6):857–879.
[45]
Houston DB (1960) Risk theory. J. Insurance 27(1):77–82.
[46]
Huber PJ (1973) Robust regression: Asymptotics, conjectures and Monte Carlo. Ann. Statist. 1(5):799–821.
[47]
Jaffee DM, Russell T (1997) Catastrophe insurance, capital markets, and uninsurable risks. J. Risk Insurance 64(2):205–230.
[48]
Jia R, Dao D, Wang B, Hubis FA, Gurel NM, Zhang BLC, Song CSD (2019a) Efficient task-specific data valuation for nearest neighbor algorithms. Proc. VLDB Endowment. 12(11):1610–1623.
[49]
Jia R, Dao D, Wang B, Hubis FA, Hynes N, Gürel NM, Li B, Zhang C, Song D, Spanos CJ (2019b) Toward efficient data valuation based on the Shapley value. Proc. Machine Learn. Res. 89(1):1167–1176.
[50]
Kamijo Y, Kongo T (2010) Axiomatization of the Shapley value using the balanced cycle contributions property. Internat. J. Game Theory 39(4):563–571.
[51]
Kashmiri S, Nicol CD, Hsu L (2017) Birds of a feather: Intra-industry spillover of the target customer data breach and the shielding role of it, marketing, and CSR. J. Acad. Marketing Sci. 45(2):208–228.
[52]
Khan F, Kim JH, Mathiassen L, Moore R (2021) Data breach management: An integrated risk model. Inform. Management 58(1):103392.
[53]
Kifer D, Machanavajjhala A (2011) No free lunch in data privacy. Proc. 2011 ACM SIGMOD Internat. Conf. Management Data, 193–204.
[54]
Kimms A, Kozeletskyi I (2016) Core-based cost allocation in the cooperative traveling salesman problem. Eur. J. Oper. Res. 248(3):910–916.
[55]
Kleinberg J, Mullainathan S, Raghavan M (2017) Inherent trade-offs in the fair determination of risk scores. Proc. Eighth Conf. Innovations Theoretical Comput. Sci., 1–23.
[56]
Kóczy LÁ (2018) Partition Function Form Games, Theory and Decision Library C, vol. 48 (Springer, New York).
[57]
Kreps R (1990) Reinsurer risk loads from marginal surplus requirements. Proc. Casualty Actuarial Soc. 77:196–203.
[58]
Krieger N (1992) Overcoming the absence of socioeconomic data in medical records: Validation and application of a census-based methodology. Amer. J. Public Health 82(5):703–710.
[59]
Krishna V, Serrano R (1996) Multilateral bargaining. Rev. Econom. Stud. 63(1):61–80.
[60]
Landes X (2015) How fair is actuarial fairness? J. Bus. Ethics 128(3):519–533.
[61]
Lange JT (1969) Application of a mathematical concept of risk to property-liability insurance ratemaking. J. Risk Insurance 36(4):383–391.
[62]
Lee YT, Sidford A (2015) Efficient inverse maintenance and faster algorithms for linear programming. 2015 IEEE 56th Annual Sympos. Foundations Comput. Sci., 230–249.
[63]
Lehtonen TK, Liukko J (2011) The forms and limits of insurance solidarity. J. Bus. Ethics 103(1):33–44.
[64]
Leng M, Luo C, Liang L (2021) Multiplayer allocations in the presence of diminishing marginal contributions: Cooperative game analysis and applications in management science. Management Sci. 67(5):2891–2903.
[65]
Maleki S, Tran-Thanh L, Hines G, Rahwan T, Rogers A (2013) Bounding the estimation error of sampling-based Shapley value approximation. Preprint, submitted June 18, https://arxiv.org/abs/1306.4265.
[66]
Mango DF (1997) An application of game theory: Property catastrophe risk load. Proc. Casualty Actuarial Soc. 84:33–49.
[67]
McQuillin B, Sugden R (2016) Backward induction foundations of the Shapley value. Econometrica 84(6):2265–2280.
[68]
Mehta S, Dawande M, Janakiraman G, Mookerjee V (2021) How to sell a data set? Pricing policies for data monetization. Inform. Systems Res. 32(4):1281–1297.
[69]
Meyers G (1996) The competitive market equilibrium risk load formula for catastrophe ratemaking. Proc. Casualty Actuarial Soc. 83:563–600.
[70]
Meyers G, Van Hoyweghen I (2018) Enacting actuarial fairness in insurance: From fair discrimination to behaviour-based fairness. Sci. Culture 27(4):413–438.
[71]
Michalak TP, Aadithya KV, Szczepanski PL, Ravindran B, Jennings NR (2013) Efficient computation of the Shapley value for game-theoretic network centrality. J. Artificial Intelligence Res. 46(1):607–650.
[72]
Michel-Kerjan EO (2010) Catastrophe economics: The national flood insurance program. J. Econom. Perspect. 24(4):165–186.
[73]
Milinski M, Sommerfeld RD, Krambeck HJ, Reed FA, Marotzke J (2008) The collective-risk social dilemma and the prevention of simulated dangerous climate change. Proc. Natl. Acad. Sci. USA 105(7):2291–2294.
[74]
Moulin H (1992) An application of the Shapley value to fair division with money. Econometrica 60(6):1331–1349.
[75]
Myerson RB (1980) Conference structures and fair allocation rules. Internat. J. Game Theory 9(3):169–182.
[76]
Nash JF (1950) The bargaining problem. Econometrica 18(1):155–162.
[77]
National Association of Insurance Commissioners (2020) NAIC announces special committee on race and insurance. Accessed January 15, 2022, https://content.naic.org/article/news_release_naic_announces_special_committee_race_and_insurance.htm.
[78]
National Research Council (2004) Measuring Racial Discrimination (National Academies Press).
[79]
Nevo A (2001) Measuring market power in the ready-to-eat cereal industry. Econometrica 69(2):307–342.
[80]
Nicholson JE, Clark K, Daraskevich G (2018) The Florida insurance market: An analysis of vulnerabilities to future hurricane losses. J. Insurance Regulation 37(3):57–89.
[81]
Nussbaumer A, Pope A, Neville K (2023) A framework for applying ethics-by-design to decision support systems for emergency management. Inform. Systems J. 33(1):34–55.
[82]
Olivieri A, Pitacco E (2015) Introduction to Insurance Mathematics: Technical and Financial Features of Risk Transfers (Springer, New York).
[83]
Pager D, Shepherd H (2008) The sociology of discrimination: Racial discrimination in employment, housing, credit, and consumer markets. Annual Rev. Sociol. 34:181–209.
[84]
Pichler A (2014) Insurance pricing under ambiguity. Eur. Actuarial J. 4(2):335–364.
[85]
Plitt S, Maldonado D (2007) Prohibiting de facto insurance redlining: Will Hurricane Katrina draw a discriminatory redline in the Gulf Coast sands prohibiting access to home ownership. Washington Lee J. Civil Rights Soc. Justice 14(2):199–254.
[86]
Rai A (2020) Explainable AI: From black box to glass box. J. Acad. Marketing Sci. 48(1):137–141.
[87]
Rockafellar RT, Uryasev S (2002) Conditional value-at-risk for general loss distributions. J. Banking Finance 26(7):1443–1471.
[88]
Roth AE (1977) The Shapley value as a Von Neumann-Morgenstern utility. Econometrica 45(3):657–664.
[89]
Roth AE, Verrecchia RE (1979) The Shapley value as applied to cost allocation: A reinterpretation. J. Accounting Res. 17(1):295–303.
[90]
Sen R, Borle S (2015) Estimating the contextual risk of data breach: An empirical approach. J. Management Inform. Systems 32(2):314–341.
[91]
Shapley L (1951) Notes on the n-person game-II: The value of an n-person game. The RAND Corporation Research Memorandum 670.
[92]
Sharapov D, Kattuman P, Rodriguez D, Velazquez FJ (2021) Using the Shapley value approach to variance decomposition in strategy research: Diversification, internationalization, and corporate group effects on affiliate profitability. Strategic Management J. 42(3):608–623.
[93]
Short ED (1985) FDIC settlement practices and the size of failed banks. Econom. Rev. Federal Reserve Bank of Dallas, March 1985, 12–20.
[94]
Shubik M (2002) Game theory and operations research: Some musings 50 years later. Oper. Res. 50(1):192–196.
[95]
Smiley KT (2020) Social inequalities in flooding inside and outside of floodplains during Hurricane Harvey. Environ. Res. Lett. 15(9):0940b3.
[96]
Squires GD (2003) Racial profiling, insurance style: Insurance redlining and the uneven development of metropolitan areas. J. Urban Affairs 25(4):391–410.
[97]
Tang S, Ghorbani A, Yamashita R, Rehman S, Dunnmon JA, Zou J, Rubin DL (2021) Data valuation for medical imaging using Shapley value and application to a large-scale chest X-ray data set. Sci. Rep. 11(1):1–9.
[98]
Tarashev N, Tsatsaronis K, Borio C (2016) Risk attribution using the Shapley value: Methodology and policy applications. Rev. Finance 20(3):1189–1213.
[99]
Thomas RG (2007) Some novel perspectives on risk classification. Geneva Papers Risk Insurance Issues Practice 32(1):105–132.
[100]
Venezian EC (1985) Ratemaking methods and profit cycles in property and liability insurance. J. Risk Insurance 52(1):477–500.
[101]
Verma V, Kawaguchi K, Lamb A, Kannala J, Solin A, Bengio Y, Lopez-Paz D (2022) Interpolation consistency training for semi-supervised learning. Neural Networks 145(1):90–106.
[102]
Wang SS, Young VR, Panjer HH (1997) Axiomatic characterization of insurance prices. Insurance Math. Econom. 21(2):173–183.
[103]
White H (1980) A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 48(4):817–838.
[104]
Xu H, Zhang N (2022) Implications of data anonymization on the statistical evidence of disparity. Management Sci. 68(4):2600–2618.
[105]
Xu Y, Armony M, Ghose A (2021) The interplay between online reviews and physician demand: An empirical investigation. Management Sci. 67(12):7344–7361.
[106]
Zhang X, Du Q, Zhang Z (2022) A theory-driven machine learning system for financial disinformation detection. Production Oper. Management 31(8):3160–3179.
[107]
Zhang X, Pérez-Stable EJ, Bourne PE, Peprah E, Duru OK, Breen N, Berrigan D, et al. (2017) Big data science: Opportunities and challenges to address minority health and health disparities in the 21st century. Ethnicity Disease 27(2):95–106.
[108]
Zheng XX, Li DF, Liu Z, Jia F, Sheu JB (2019) Coordinating a closed-loop supply chain with fairness concerns through variable-weighted Shapley values. Transportation Res. Part E Logist. Transportation Rev. 126:227–253.
[109]
Zhou WX, Bose K, Fan J, Liu H (2018) A new perspective on robust m-estimation: Finite sample theory and applications to dependence-adjusted multiple testing. Ann. Statist. 46(5):1904–1931.
Recommendations
Risk Assessments and Fairness Under Missingness and Confounding
AIES '19: Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and SocietyFairness in machine learning has become a significant area of research as risk assessments and other algorithmic decision-making systems are increasingly used in high-stakes applications such as criminal justice, consumer lending, and child welfare ...
Catastrophe risk modeling and application-risk assessment for Taiwan residential earthquake insurance pool
MS'06: Proceedings of the 17th IASTED international conference on Modelling and simulationTaiwan lies at earthquake-prone area. More than 200 sensible earthquakes occurred every year in Taiwan. Average annual loss due to 83 disastrous earthquakes since 1900 is about NT$19 billion dollars which equals to 0.7% GDP. Therefore, the Taiwan ...
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
Copyright © 2023, INFORMS.
Publisher
INFORMS
Linthicum, MD, United States
Publication History
Published: 17 January 2023
Accepted: 06 December 2022
Received: 15 February 2022
Author Tags
Qualifiers
- Research-article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 03 Oct 2024
Other Metrics
Citations
View Options
View options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in