Does Big Data Discriminate?

Are Big Data practices discriminatory or not? Does data help or hurt underserved communities?

These are the questions many in Washington are asking. The Federal Trade Commission just announced a workshop on whether Big Data is a “tool for exclusion or inclusion.” The Department of Commerce is jumping on boardtoo. Meanwhile, the White House issued a report a few months back detailing the “potential for big data analytics to lead to discriminatory outcomes.”

We should be asking questions about Big Data, but let’s be clear about the real problems and real solutions.

Big Data is just a tool—it can be used for good or ill. What we do know is that Big Data itself is not the enemy. Simply singling out one innovation or industry won’t rid us of discrimination. Prejudice has left a longer, sadder mark on human history than today’s technologies. 

We need to focus on the good uses of data while substantively addressing those practices that are clearly discriminatory or abusive. Achieving these aims require more data, not less.

Among the positive uses of Big Data are its potential for empowerment and inclusion. Philanthropies and companies are using data-driven insights to deliver betters outcomes in health and welfare for underserved communities across the world. 

DataKind, for instance, linked up their volunteer data scientists with the Grameen Foundation to determine what information Ugandan farmers needed and when. Kaiser Permanente is looking at how wearable devices and improved analytics could empower families to better take care of themselves, while quickly alerting medical professionals when an emergency has occurred. With these and many other examples, knowledge is power.

Civil rights groups stress the ways in which data collection can be used to identify harmful practices, and advance anyone who’s been trodden under foot through no fault of their own. Better access to and sharing of data empowers individuals and communities to engage in a more open marketplace, whether for banking services or healthcare or starting a business.

Just as significant is the ability of data-driven innovation to help us lead our lives more efficiently and with greater convenience. After all, life for the poor can be a constant series of stop signs and road blocks that feel purpose-built to slow their advancement up the economic ladder. Those charged with fighting poverty can now know which areas are most in need of their help, and to do so with better delivery and design.

Companies and nonprofits can also use data in harmful ways—that much is true. We should identify what those uses are, whether in adversely targeting individuals or engaging in deceptive practices, and place them firmly off-limits.  

By identifying clear “databuses,” we can preserve every other use of data that enables innovation and benefits consumers. From these we should derive best practices that can be measures by which we judge good behavior in the private sector. Doing so requires more data, not less.

When it comes to real concerns about data discrimination, we should focus on empowerment, opportunity, and identifying harm. “Data for good”—that should be our motto.