In chapter 9 of the book Weapons of Math Destruction, the author discusses about how big data influences people's ability to get any kind of insurance. At first, we learn about how statisticians started predicting certain life events and based on that they came up with the idea of insurances which eventually turned out to be a profitable business niche. Insurance is meant to provide assurance to individuals in distress. However, with time insurance companies have found various ways to exploit individuals with high premiums and incentivized people to give up their privacies minimizing risk and optimizing revenue. And, big data has played a key part in how the insurance companies started learning about consumer behavior to eventually impose controlling measures over their life.
As with any other WMD, racial stereotypes were one of the most favorite attributes of the statisticians. Such stereotypes are also factored in to consider an individual's insurability. However, due to unstratified data, most statisticians end up redlining certain neighborhoods and groups of people without even considering their socio-economic background. Some auto insurance companies use credit reports for determining one's insurance capacity. Using financial metrics to determine one's behavior on the road is heavily misleading. Eventually, when people started raising their voice to convince auto insurance entities to use one's domain specific data to determine insurability, the insurance companies started capitalizing on these very domain specific data. They found ways to gather granular data on consumer behavior which was beneficial for some but biased for most in the sense that these data started bucketing people on how they act. So, even if someone is quite eligible for insurance savings, she might still be paying the high premium for being in a risky bucket.
Insurance misery has not spared the employment sector as well. Employee surveillance has keyed into optimizing revenue from employee paychecks through high insurance premiums. Employers have introduced corporate wellness programs that coerce employees to follow certain health measures to reduce their insurance cost. Such programs are definitely well-intended. But, withholding insurance rights against enrollment in such programs stirs controversy. On top of that, data suggests that such wellness programs don't really improve employee lifestyle or productivity in the long run.
In both the auto and health insurance industry, personal trackers are being introduced by incentivizing good (read expected) behavior. As a result, lives of people are being modeled by WMDs that are based on these tracker data. If individual's wish to opt out of such tracking based reward policies, they are implicitly forced to pay a higher premium. Evidently, privacy has now become a luxury for people. Only a handful willing to pay the extra are possibly being spared of such intervention. So, an obvious way to solve this issue will be respect privacy and design such tracking programs as opt-in only without incentivizing them in general. A person will want to enroll in such a program not because she might be able to save some insurance premium from it, but because she has intentions to improve her lifestyle.
