By now, you’ve heard of big data – the large-scale volume of information collected from everyday actions and behaviors. You’re contributing data to this web of information every time you post on social media or swipe your credit card. But do you know how that big data you’re creating is used?
Big Data Used for Discrimination in China
China is planning to score all its citizens on a variety of behaviors it deems positive. While only pilot projects are underway now, the Chinese government wants to create a database that combines data about citizens and businesses in one place. Any interaction — in the real world or online — could be captured and stored, along with information that could produce a score based on their political, commercial, social and legal “credit,” according to The Independent.
Comments on social media, online shopping and content consumption are only the beginning. The system could also combine records from courts, the police, banking information, tax payments and employment records. Moreover, professionals and companies might also be scored for professionalism and honesty.
Proposed punishments for low social-credit scores include not being able to take first-class seats in planes and trains, booking in luxury hotels, and sending children to the best schools.
A Chinese policy document released this fall warned, “If trust is broken in one place, restrictions are imposed everywhere.”
Big Data Harms the Little Guy
Big data may also keep the poor down. Most consumers are highly aware of their FICO scores, better known as “credit ratings”, that are created by taking into account a person’s history of payments. While FICO scores are at least objective and transparent, they are being used to make decisions outside of providing credit. For example, some companies won’t hire people with bad credit, so having a low credit score can prevent someone from getting a needed job, putting that person into a spiral of debt — and worse credit scores.
Now, a new generation of companies is scoring consumers on hundreds or thousands of individual data points beyond whether they pay their bills on time — criteria that seem not so far removed from those Chinese citizenship scores. They’re variously known as e-scores, electronic credit scores or consumer value scores. One such provider, eBureau, says its scores can be used for marketing, lead management, fraud prevention, credit assessments and collection decisions.
We may be rated on how well we adhere to a doctor’s prescription regimen, our health risk, how likely we are to remain at a job, how likely we are to engage in fraud, to gamble and how profitable we might be to a retailer, according to the World Privacy Forum. Unlike credit scores, which are based on objective data, consumer scores are based on secret criteria — so they can hide discrimination and bias, the forum said.
People with lower consumer scores may have less access to consumer credit, loans and mortgages. They may not be shown ads offering discounts or special deals. They may not receive the same level of customer service as those with “better” scores do.
The New York Times called these consumer scores “a private, digital ranking of American society unlike anything that has come before.”
Cathy O’Neill, author of Weapons of Math Destruction, told NPR that big data used this way is especially dangerous when it has a high impact on people’s lives, such as determining whether or not someone gets a job, when the criteria are secret, and when it harms someone’s life.
Data to Reduce Bias
Big data can also be a tool for inclusion, according to a report from the Federal Trade Commission.
Virginia Commonwealth University is using big data to help identify students who are at risk of dropping out and in need of early intervention strategies, according to the Washington Post. Counselors may help such students reevaluate their majors or make sure they register for the right sequence of classes. In the first semester alone, the school recorded a 16 percent increase in the number of students completing courses and an 8 percent increase in students enrolling for the following term.