The conventional answer — of course not! — animated public outrage last year after Facebook published a study in which it manipulated how much emotional content more than half a million of its users saw. Similar indignation followed the revelation by the dating site OkCupid that, as an experiment, it briefly told some pairs of users that they were good matches when its algorithm had predicted otherwise.
But this outrage is misguided. Indeed, we believe that it is based on a kind of moral illusion.
Companies — and other powerful actors, including lawmakers, educators and doctors — “experiment” on us without our consent every time they implement a new policy, practice or product without knowing its consequences. When Facebook started, it created a radical new way for people to share emotionally laden information, with unknown effects on their moods. And when OkCupid started, it advised users to go on dates based on an algorithm without knowing whether it worked.
Why does one “experiment” (i.e., introducing a new product) fail to raise ethical concerns, whereas a true scientific experiment (i.e., introducing a variation of the product to determine the comparative safety or efficacy of the original) sets off ethical alarms?
In a forthcoming article in the Colorado Technology Law Journal, one of us (Professor Meyer) calls this the “A/B illusion” — the human tendency to focus on the risk, uncertainty and power asymmetries of running a test that compares A to B, while ignoring those factors when A is simply imposed by itself.
Consider a hypothetical example. A chief executive is concerned that her employees are taking insufficient advantage of the company’s policy of matching contributions to retirement savings accounts. She suspects that telling her workers how many others their age are making the maximum contribution would nudge them to save more, so she includes this information in personalized letters to them.
If contributions go up, maybe the new policy worked. But perhaps contributions would have gone up anyhow (say, because of an improving economy). If contributions go down, it might be because the policy failed. Or perhaps a declining economy is to blame, and contributions would have gone down even more without the letter.
You can’t answer these questions without doing a true scientific experiment — in technology jargon, an “A/B test.” The company could randomly assign its employees to receive either the old enrollment packet or the new one that includes the peer contribution information, and then statistically compare the two groups of employees to see which saved more.
Let’s be clear: This is experimenting on people without their consent, and the absence of consent is essential to the validity of the entire endeavor. If the C.E.O. were to tell the workers that they had been randomly assigned to receive one of two different letters, and why, that information would be likely to distort their choices.
Our chief executive isn’t so hypothetical. Economists do help corporations run such experiments, but many managers chafe at debriefing their employees afterward, fearing that they will be outraged that they were experimented on without their consent. A company’s unwillingness to debrief, in turn, can be a deal-breaker for the ethics boards that authorize research. So those C.E.O.s do what powerful people usually do: Pick the policy that their intuition tells them will work best, and apply it to everyone….(More)”