What do you think? Leave a respectful comment.

Lifestyle choices could raise your health insurance rates

Health insurance providers are paying data brokers to find personal information about their clients -- race, marital status, ability to pay bills and more -- to predict client health costs. But unlike medical records, there aren’t any laws that regulate how insurers can use the information. ProPublica reporter Marshall Allen, who wrote a series on the issue with NPR, joins Hari Sreenivasan.

Read the Full Transcript

  • HARI SREENIVASAN:

    What you pay for health insurance is increasingly a complex web of formulas. And now your personal data — everything from where you live, to what clothing you buy, to your magazine subscriptions — may factor into what you pay or whether you get coverage at all. In a series of reports published with NPR, the investigative nonprofit news organization ProPublica is looking into the strategies insurance companies are using. And joining us now from Denver is ProPublica reporter Marshall Allen. Marshall, first, what sorts of data are they looking at and what sort of inferences can they make from that?

  • MARSHALL ALLEN:

    Well that’s a good question, Hari. They’re looking at all different types of personal and proprietary and public information. The kinds of things that people would normally assume to be private. And I bet probably most of the viewers in your audience right now are having their data gathered by the data brokers that are teaming up with the health insurance companies to analyze this. And so the data that they’re gathering would include your education record, your property records, any debts you might have, your income level, your race and ethnicity, even social media interactions, they’re gathering those. So, they’re gathering all this information and they’re putting it into these complex computer algorithms and then they’re spitting out predictions about how much we might cost based on all these economic and lifestyle attributes.

  • HARI SREENIVASAN:

    So give me some examples.

  • MARSHALL ALLEN:

    Well so, for example, some of the inferences they make are creepy, I guess. You know, you could say this kind of turns the creepiness level up a bit. For example, they can tell if a woman has changed her name. And so they say if a woman has changed her name in the last 24 months, she may have recently gotten married, or maybe she recently got divorced, and so she could be considering getting pregnant soon or maybe she’s stressed from that divorce and so she’s going to have a lot of mental health care costs. Or if you’re a low income minority, they would assume that you are living in a dangerous and dilapidated neighborhood. And so you could be at higher risk of health cost. Or another one is if a woman has bought plus sized clothing, they would predict that she might be more likely to be depressed which could also lead to higher health care costs. So these are things that they’re looking at: trends in the data for groups of people, and then they’re attributing the inferences to individuals within that group, and one of the fundamental problems is that for any individual, this could just be wrong.

  • HARI SREENIVASAN:

    And so what if they’re wrong about this? You’re still going to be falling into a bracket based on this group and the suspicion that they have that you are part of it?

  • MARSHALL ALLEN:

    Yeah, I mean, they’re scoring us and predicting our health care costs based on the groups that we fall into. And so I went to LexisNexis and obtained, they’ll give you a certain portion of your data, and it was like a creepy walk down memory lane for me. They had data for me going back 25 years to the address of the home I grew up in in Golden Colorado. You know, all my old phone numbers. And with each of the addresses, you know, they had a little indicator there: was this a high risk neighborhood or not? And I grew up in a middle class kind of environment, so I didn’t grow up in any high risk neighborhoods. But it made me wonder, what if I had? And when I talk to the industry, I mean, they promise that they’re only using this information for the purpose of helping people. So, what they would say, their argument for doing this is that they can do better case management. But just as it could be used for good, it could also be used to discriminate, and the health insurance industry has a long history of discriminating against sick people. That still goes on to this day.

  • HARI SREENIVASAN:

    Marshall Allen of ProPublica joining us from Denver today. It is part of a yearlong reporting project called the Health Insurance Hustle. You can find it on their website. Thanks for joining us today.

  • MARSHALL ALLEN:

    Thank you, Hari.

Listen to this Segment

The Latest