98295 - SOCIAL CONSEQUENCES OF ALGORITHMIC PREDICTION

Learning outcomes

The aim of this course is to provide an overview of the increasingly spreading of data-driven automated decision-making in social systems, with special focus on insurance practice. At the end of the course the student will be able to: • assess social issues of fairness and discrimination in algorithmic forms of prediction • explain social assumptions and consequences of predictive analysis in insurance societies • analyze and interpret concrete cases of data-driven decisions in the field of social and private insurance.

Course contents

Insurance is a pillar of modern society. Its business model is based on statistical and probabilistic calculation. The algorithmic turn of prediction promises to disrupt this business model with far-reaching social consequences. The course will focus on the impact of usage-based insurance policies on society as a whole. The first part of the course deals with the main mechanism underlying the insurance business, its preconditions and sociological relevance. In the second part of the course we discuss the main consequences of behavioural data-driven predictions and the changing idea of prevention in modern social systems.

Against the backdrop of the topic of usage-based insurance, course participants will familiarize with the issues of behavioural valuation of policyholders in contemporary society. Within the framework of a sociological theory of insurance, they will acquire the skills required to observe and critically comment on the diffusion of digital innovations in the insurance industry. The focus on InsurTech and usage-based insurance policies will allow participants to concretely examine some crucial issues such as fairness, information asymmetry, preventive strategies, mutuality and risk-transfer mechanisms.

Readings/Bibliography

A detailed syllabus will be provided at the beginning of the course.

Teaching methods

The course is organized according to the model of the Structured Seminar. Twelve hours will be taught in six 2-hour classes (first three lectures weeks) that introduce the basic concepts and crucial issues of insurance as a social institution based on prediction. The remaining classes will be organized as seminars held in presence in fourteen meetings (2 hours each).

Participants are expected to read the materials in advance and actively contribute to the discussion. They are supposed to present at least a paper and prepare two questions for each meeting, that can be presented and debated during the sessions. After each session, they will deliver a short memo (e.g. several bullet points) summarizing their understanding of the outcome of the meeting in terms of the issues they find most relevant. At the end of the course, they will write a paper commenting and discussing one of the topic discussed during the meetings.

Assessment methods

Class attendance: 35%

Oral presentation: 25%

Written memos and paper: 40%

Office hours

See the website of Alberto Cevolini