Discrimination is an ongoing problem in the insurance industry that persists, regardless of intentionality, when the insurer blinds the pricing process from the socially controversial or legally prohibited input. We contextualize the problem in P&C insurance considering the prevailing legislations in the United States and the European Union.

We first analyze the largest publicly available database of police-reported motor vehicle traffic accidents in the United States. Then, we set the pricing problem under a causal framework to provide tools that can help identify sources of discrimination. This leads to a criteria we propose to obtain actuarially fair premiums that can mitigate discrimination. The aforementioned comes together in a microsimulation approach for auto insurance in the U.S. that approximates population statistics and reported accident data reported. Our results show that some traditional actuarial assumptions can result in unintended discriminatory consequences.