Big Data or Big Brother?
Big data has limitless potential. We are already using it to do everything from scientific breakthroughs such as decoding DNA strands and predicting disease patterns, to more trivial every-day conveniences such as suggesting what TV programmes we might be interested in watching. Many of the world’s leading organisations are leveraging big data to take a strategic analytics approach to achieve sustainable marketplace advantage. However, there are a lot of gray areas when it comes to the ethics of data collection and analysis. The rapid increase of data mining in corporate America has brought about much debate around the intentions and ethicality behind it, most recently around health.
Employers in America are using wellness firms including Castlight Health and Welltok to use big data as a means of predicting employee health needs. Typically, these firms gather information about the prescription drugs employees use from insurers and predict which illnesses they might be at risk for. Some actually buy information from data brokers that help to connect people’s consumer spending habits with their health care usage. In fact, it’s been said that where you shop and where you eat is a better prediction heart attack risk that your genome.
Supposedly, the intention is to help employees to improve their own health, lead allround healthier lives and find lower cost healthcare options, but at the same time, slashing corporate healthcare costs. Employees that are identified as being at risk for certain conditions can opt in to directly receive emails from the health data firm with care tips, recommendations to doctors, and alternative treatment options. Taking spinal surgery as an example – a procedure that costs upwards of $20,000 – an algorithm would be able to identify employees who might be considering costly spinal surgery and send them information about cheaper fixes or alternative treatments such as physiotherapy.
All of this data is anonymised and aggregated so employers will only see the number of people who are at risk and they won’t receive data for groups smaller than forty people. It has been stressed that managers cannot get information about individual employees. In saying that though realistically only a random sampling of an employee population is needed to extract meaningful and actionable data, so the data can still be used to reprimand employees who did not opt in.
Castlight recently launched Castlight Action – a product that has the ability to determine the likelihood of a woman being pregnant. It scans insurance claims to identify women who have stopped birth-control prescriptions, as well as women who have made fertility-related searches on the Castlight health app. It then cross checks this with a number of variables including age and number of children and calculates the likelihood that she is pregnant. They will then receive information on obstetricians and staying healthy throughout pregnancy but many women are likely to find this unsettling and intrusive in a work environment.
Castlight Action resurfaces bad memories of when Target let a teenager’s father know about her pregnancy before she got the chance to. Target sent coupons for baby items to customers according to their pregnancy scores and shopping habits but failed to take age into account. The teenager’s father deemed the coupons to be inappropriate and encouraging teenage pregnancies, and so made a complaint, only to learn later that his daughter was, in fact, pregnant. Target had to learn the hard way that it sometimes makes sense to pretend like you know less about your customers than you actually do, to avoid creeping them out or invading their privacy.
The area of big data remains somewhat unregulated. At present, there are almost no laws that control what data these big data companies can access and little law controlling what they can do with it. However, with such sensitive data comes serious security concerns, hacks and breaches and when it involves people’s behaviour patterns and how they live, it becomes more of a moral issue. Additionally, there is the frightening thought that employers might find out about a pregnancy or illness before the employee is ready to disclose that information and that brings up many other issues.
The topic of big data is enormous; a vast expanse of issues that are both murky and unchartered. Addressing it runs the risk of biting off more than you can chew, much like the exploit of the phenomenon itself.