With just three commonly available data points — zip code, gender, and date of birth — 87 percent of Americans can now be uniquely identified. With the massive stores of data compiled by major aggregators like Google, Facebook, LinkedIn, and Yelp, that number — and the level of detail that can be obtained — skyrockets.
This trove of consumer data has been a boon to business, allowing companies to target the right offer to the right consumer at the right time. But, as Bruce Kogut points out, data science also presents businesses with major ethical quandaries. “If I predict someone’s credit worthiness by where they live, that’s illegal,” he says. “But if I use their Facebook data and predict their creditworthiness based on who their friends are, that’s legal. But that’s the same thing, right?”
While much of the public debate on privacy and information collection has focused on consumer data like this, businesses increasingly collect similar information on the behavior and communications of employees. That shift, according to Kogut, represents a fundamental change in the nature of the workplace, leading him to assert, “data science will make or break our world.”
Concerns over privacy, data mining, and the use of algorithms — increasingly necessary to sort through the mountains of collected data — can only be amplified when the use of data turns from serving ads to determining fundamentals like who gets jobs and how much they’re paid. Rather than correcting for human biases, algorithms might further exaggerate them. Noting, for example, that gender might predict success in an organization, an algorithm, mistaking correlation for causation, may suggest more male candidates for promotion, increasing gender imbalances in leadership. “A good predictor isn’t necessarily a good causal identifier,” Kogut points out, “and is often neither true nor fair.”
Even when these programs work as they’re supposed to, however, they can still create problems. “Going back to the 1800s,” Kogut says, “the thinking has been that people who are highly productive should be matched with higher-paying jobs, and those who are less productive should be placed into lowerpaying jobs. But, that’s a very brutal world.”
Worse, it may be bad for business. In a recent study with Jerry Kim, Kogut found that a retail chain that took a data-driven approach to identifying high performers and rewarded them with valued benefits, like scheduling priority, experienced higher employee turn over, at substantial cost and without any increase in productivity. By placing higher pressure on individuals for short-term growth, these policies can turn co-workers into competitors, impeding information sharing and learning and ultimately handicapping productivity growth in the long run.
If today’s unprecedented access to troves of data presents us with ethical quandaries, however, it might also provide solutions. “Because of these tools, we’re actually now able to do what I call forensic ethics,” Kogut says. “We can go in, take large amounts of data from a company, and we can very quickly isolate abnormalities that suggest ethical problems.”
By utilizing what data teaches us about individual workers and the connections between wages, productivity, and performance, Kogut believes we can foster greater collaboration within organizations while promoting diversity and greater pay equality, which will be as good for business as they are for society. What is needed is not less data, he suggests, but richer data, and a more sophisticated approach to the challenges it presents.
It’s a critical time for action, he says, as the workplace is undergoing a “dramatic change,” led by shifts in the economy at large. “Is this sharing economy a good thing or a bad thing?” Kogut asks. “Is Uber good or bad for drivers? We need to analyze these questions and how they affect other problems like the wage stagnation for middle class workers over the last 20 or 30 years.”
This essay originally appeared in the centennial edition of Ideas at Work. View the entire collection online here.
About the researcher
Bruce Kogut is the Sanford C. Bernstein & Co. Professor of Leadership and Ethics at Columbia Business School. He teaches courses on Governance and...Read more.