Actuaries are the mathematicians for insurance companies, says Dorothy Andrews, the chief behavioral data scientist for the Actuarial and Analytics Consortium.
As big data expands what information insurers factor into your risk profile, their calculus is more precise than ever. Andrews believes insurers should provide policyholders transparency about that information and how it’s being used. When she thinks What The Future, she sees a world where there are no secrets and that’s a good thing.
WTF: Can big data lead to fairer insurance rates?
Andrews: The greater the risk profile of a person, the higher their risk premium will likely be. The less risk, ideally, you get your insurance at a lower rate. Having more information is really intended to make that pricing equation more informed and risk appropriate.
WTF: What would having more access to consumer data do for the insurance industry?
Andrews: When insurers know as much about you as you do, they can develop a more risk-appropriate price. In the process, you can potentially differentiate more profitable customers from less profitable ones. The problem with that approach, of course, is that insurance can also become unaffordable for people who need it. So, there’s going to have to be some kind of societal mechanism to make that whole equation work.
WTF: What kind of information is being pulled from these newer big data sources?
Andrews: There are UBI [usage-based insurance] devices from an insurance company that you can put in your car that will collect your driving behavior data. You probably have a tracker that you wear that keeps track of your steps and your sleep patterns. There are continuous glucose monitors for diabetics in addition to lots of other monitors out there for other kinds of biometrics. We have smart home devices that collect data. We’re becoming surveilled and quantified in just about every aspect of our lives.
WTF: Is that the future of insurance you expect?
Andrews: As we become more self-quantified—through self-tracking devices, we are enabled to more efficiently manage our lives. However, the companies enabling you to see your data are also aggregating the collected data over many lives and monetizing it. As someone who is becoming a media psychologist, I have a problem with that. Mainly because the people whose data is being aggregated are not really being compensated for their information being used in this profit-seeking kind of way.
WTF: Should insurers be required to detail the personal data that they use to determine their premiums and coverage?
Andrews: Oh, absolutely. Not only should they tell you what they’re using. They should tell you how valid it is. They should make sure they share it with you so you can determine and assess whether it’s error-free. The age of the data they are using should be current to the applicable insurance coverage period. Sometimes you will find companies using data that’s more than 20 years old. How does that data relate to you as a risk now? Only relevant, current data should be used.
WTF: That’s a really great point. Our survey shows that most people agree that wearable technology can improve your health. But most people feel like the data about them is going to give them a less fair price.
Andrews: I don’t like the word “fair.” It is so hard to define. Sometimes people may tend to think, for example, “If the insurance company doesn’t know I have a heart problem, then they can’t penalize me for it.” Now that we live in the age of big data, correlation has become more important than causation. If someone has heart disease, there’s probably other data that’s correlated with having heart disease that would help the insurer detect the condition even if it is not disclosed.
WTF: Could insurers create incentives or penalties for addressing those gaps?
Andrews: Yes, but we have to stop treating adverse information about a person in such a way that they are less likely to disclose it. Some companies have put programs in place to help improve policyholder’s risk profiles after the policy is issued. Such programs are designed to move people toward a standard (or better) risk profile which will reduce their premiums.
WTF: We spoke with Dr. Aletha Maybank at the American Medical Association for our gender issue about health equity including how people want to see physicians that look like them. Does that apply to insurance?
Andrews: Yes. In fact, this is not a trivial topic. Media psychologists and other social scientists are concerned with the lack of diversity among programmers of A.I. because of the potential to introduce biases into the programs. It’s really important for people who are building insurance models to also reflect diversity so that mistakes of the past made in other disciplines are not repeated in insurance risk modeling.
WTF: One concern in the insurance industry is that big data and AI could put actuarial professionals out of business. What do you think?
Actuaries have a skill set that can metamorphose as technology changes and becomes more pervasive. In my opinion, we are going to be just as valuable, if not more.
WTF: By 2040 do you think we’ll still have questions about people either hiding or withholding their personal data from insurers?
Andrews: If my vision becomes reality, there are not going to be any more secrets. That’s how we’re going to drive efficiency. As we start to change our cultural mindset about data being available and not feeling that we have to be so protective about it, and most importantly, make it useless in the wrong hands, there will be no limits to how we can exploit technology to improve the human condition. Let’s say, as an oversimplified example, I knew your Social Security Number and there was absolutely nothing I could do with it. Would you really care if I knew it? This is something we need to start working on because we really can’t advance a lot of technology without having access to data.