In 2012, Katherine Milkman, a professor at Wharton who studies judgment and decision-making, co-authored a study that sought to determine the role of race and gender in professional advancement. In order to do that, Milkman and her colleagues used 20 names that might be associated with a particular race or gender and assigned them to fictional prospective doctoral students. The researchers used those names to send identical emails to more than 6,000 professors asking to discuss their research and doctoral programs. The responses varied significantly based on the presumed race of the fictional candidate, and the study helped spark a public conversation about discrimination in academia.
Since then Milkman has continued to study the ways people make decisions, and how those processes can be altered to promote equality. Her work has also prioritized exploring the kinds of biased decision-making that leads to the underrepresentation of people of color and women that plagues many professional fields.
I spoke to Milkman for The Atlantic’s series on mentorship, “On The Shoulders Of Giants,” about her research, and the counterintuitive ways that everyday networking can harm women and minorities. The interview that follows has been edited for length and clarity.
O’Donnell: What biases have you found to be most harmful to mentorship?
Milkman: Stereotyping, where we assume that people having certain characteristics—how they look, how they talk—allows you to make assumptions about their abilities, or their interests. Even if you have the best of intentions as a mentor, you may inadvertently use stereotypes to shape the kinds of advice and pointers that you offer to your mentees.
O’Donnell: Can you think of an example?
Milkman: For instance, if you have multiple mentees, and you think an opportunity might be beneficial for one of them, you should stop and say, “Hey, could this be beneficial for the others? Is there a reason I'm not thinking of them? Is there a true distinction in their interests or what would benefit them? Or could it possibly be stereotyping?”
O’Donnell: Your work on decision-making bias in mentorship has expanded on the concept of “micro-inequities.” What are those and what role do they play in people’s professional lives?
Milkman: Small things add up over the course of a career, including a lack of opportunity, or a lack of networking. In our work, we looked specifically at the encouragement that prospective doctoral students received from faculty when they approached them about their research and the possibility of working with them. If a professor’s decision to respond or not is based on any kind of stereotyping, we argue that this is exactly the type of small act that, when aggregated, can have a large impact on a person’s career. If every time you seek encouragement you face these inequalities—less encouragement, less responsiveness—it can add up to substantially fewer women and minorities in the pipeline.
O’Donnell: What is the most counterintuitive thing that behavioral economics has to say about how bias can impact people’s careers over time?
Milkman: In the context of networking, which is very relevant to mentoring, we are likely to try and help people in our social networks by making connections for them. I might say, “Oh, you’re interested in journalism—I was just talking to a journalist that writes for The Atlantic.” We proactively or even just reactively offer helpful connections, guidance, and opportunities for people in our social networks, because that's part of being a good person, frankly.
The thing is, every time we offer help to one person, that’s actually harming all the people who aren’t receiving help. If one person gets a leg up, that’s a leg down for whomever else is competing for those opportunities.
O’Donnell: And what does that mean for women and people of color?
Milkman: White men tend to have the best social networks and be the best connected—that is an empirical fact. What that means is that when we are all going around in our daily lives, trying to help the people we know, the act of helping those people disadvantages women and minorities—because at scale white men preferentially get those networking advantages. And so that’s another really pernicious bias that I think as a mentor we have to be very aware of: Who are we helping?
O’Donnell: I’m guessing that the takeaway from this research isn’t to stop helping people in your network. But what can be done to balance these issues?
Milkman: It’s a really terrific question and one I struggle with when I am teaching this in front of a whole classroom of students who ended up in my MBA classroom through their networks. A huge part of their lives is networking. They sort of gasp in awe that that could be somehow a bad thing. I think the most practical prescription is not to stop helping the people in your network and start being a jerk to everyone.
O’Donnell: So what is the most practical prescription?
Milkman: I think that the best advice is to realize that, by helping your network, you are hurting other groups. Rather than feeling like it’s crazy to acknowledge that two people with slightly different resumes might have had slightly different opportunities, maybe you need to recognize that even if their CVs look slightly different, maybe the person who had fewer opportunities is actually going to be a better candidate.
Whenever you do have someone in your network from a group that is underrepresented—a woman, a minority, a first-generation college student—we should be more aggressive about helping them. Recognize that it’s even more important to do things for them—maybe do three times as many things as you usually would for those individuals. I think that’s the best prescription.