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.
The early days of the SARS-CoV-2 (or COVID-19) pandemic have seen a proliferation of calls for “individual responsibility”, starting with strong calls (and even an obligation in some countries, including France) to stay home as much as possible in the early spring of 2020, before it became mandatory to wear a mask in public (often closed) places during the summer. To paraphrase Coluche « dire qu’il suffirait que les gens restent chez eux pour qu’on puisse sortir… ». This call for each person’s responsibility is made in the name of all and for the good of all, symbolizing this very particular solidarity that the pandemic reminds us of: the risk that I choose to run does not only concern my person but also constitutes a risk for those around me. To formulate it in probabilistic terms, McKendrick (1926) stated that “the probability of occurrence increases with the number of existing cases“. This conception of individual responsibility, which is quite intuitive a priori, actually runs counter to the classical conception of economics: the rational (and responsible) individual makes choices that concern him, and that concern only him. The collective good is deduced by summing up individual utilities, independent of each other. But here is the problem: with the epidemic, an interdependence of utilities is created, so that the well-being of so-and-so, who chooses not to wear a mask, can harm the health and therefore the utility of many other people. How then can we think in economic terms of this “individual responsibility” in the context of the epidemic?