“Progressive” Auto Insurance
What if I told you that you might be paying up to 47% more for auto insurance simply because you’re renting a house in Ann Arbor? I would be pretty upset too. Did you know you could be paying up to 11% more if you’re single? The number goes up if you’re working in the service industry, and even if you’re unemployed. How does this make any sense? One would think that your prices would go down if you don’t have a job — no job to drive to means less time on the road means lower chances for an accident. Or at least one would think.
Insurance agencies are not immune to the powers of weapons of math destruction either. If anything, they only benefit from the discriminatory tendencies inherent to their algorithms. This has been a problem for a long time. With articles dating back to 2012, insurance companies have been profiting off of marginalizing specific demographics for at least a decade now. These habits are both racist and sexist. The data used to determine these price hikes are often proxies for socio-economic factors which do not reflect driving habits, but rather race and sex.
“These mathematical models… even when wrong or harmful, were beyond dispute. And they tended to punish the poor and oppressed in our society.” (O’Neil 2016) It’s apparent that insurance companies are simply throwing a bunch of numbers at an algorithm and letting the computers decide for them, otherwise factors like unemployment wouldn’t incur additional penalties. I would go so far as to say that the unemployed are the ones who require discounts given their circumstances as well.
“The bottom line for numbers is that they cannot speak for themselves…. Especially when the numbers have to do with human beings, they run the risk not only of being arrogantly grandiose and empirically wrong, but also doing real harm in their reinforcement of an unjust status quo.” (D’ignazio and Klein 2020) Especially within state mandated auto insurance, the recognition of context of data factors needs to be thoroughly examined and respected before slapping it into an algorithm. In this case, lower income drivers, African Americans, and women are unfairly given higher rates because the mathematicians behind it neglected to consider the implications of the factors regarding driving ability, and inadvertently created an entire industry that is systemically racist and sexist. In the right hands, auto insurance could both remain profitable and give reasonable rates directly correlated to driving ability, and potentially even help those who are economically disadvantaged.
