Professor of Public Policy Sharad Goel looks at public policy through the lens of computer science, aiming a computational and statistical lens at a diverse range of contemporary social and political issues including criminal justice reform, democratic governance, and the equitable design of algorithms. Prior to joining Harvard Kennedy School, Goel was an assistant professor at Stanford University and he holds a master’s degree in computer science and a PhD in applied mathematics from Cornell University. He teaches DPI-617, “Law, Order and Algorithms.”
Q: Policy is becoming increasingly intertwined with technology. What drew you to study and teach about the place where they intersect?
About ten years ago, I was living in New York City and there was a high-profile federal class action lawsuit going on involving stop and frisk. There was a line in the judicial opinion that struck me. It said that all of these people’s rights were likely being violated, but we’ll never know how many. I got involved in the issue from the sidelines and started developing empirical methods for understanding the scale of these constitutional violations.
I ended up in New York after finishing my PhD at Cornell and a post-doc at Stanford. I was doing pure math at the time, and I just kind of burned out and wanted to try something new, so I went to New York. The stop-and-frisk project was only tangentially related to my day job at Microsoft Research. After I started doing more of this kind of public interest-oriented work, I really began to see the power of computational and statistical methods for addressing problems that were very much top of mind for me and for many people. That project got me excited about the possibilities.
Q: You're an applied mathematician and you often write about policy issues from that statistical perspective. Why do you think that approach is important and effective?
I don’t think it’s the only approach, but I do think it’s complementary to other approaches. Maybe because of my training, this is just how I view the world, through this statistical, computational lens. And for me, at least, that vantage point helps highlight some of the subtleties in complex social problems. I find that data are often useful for untangling nuanced policy questions—including understanding the nature of discrimination and ways we can address it.
At the same time, I think taking a clinical approach to these problems is not always the right one, or the most effective one, or the one that best resonates with people. So I personally try to strike a balance, talking to both policymakers and folks who are directly impacted by the issues we study, working to understand where they are coming from, and coupling their perspectives with the statistical analysis.
“Data is often useful for untangling nuanced policy questions—even for understanding the nature of discrimination and ways we can address it.”
Q: You teach a class called “Law, Order and Algorithms,” which is interesting because it seems technology is proving to be the ultimate double-edged sword. It is helping us discover previously submerged biases and inequities, but it's also driving them, and it’s increasingly seen by many people as a threat to everything from societal function to democracy. What do you think can be done to push this duality of technology in a positive direction?
These are tough questions! I think that tension won’t go away. Technology enables the creation of powerful tools, which can both improve social systems and exacerbate inequities. It’s important for us to help current and future policymakers—including our students—become better versed in the value and limitations of technology. With that knowledge, they'll be better equipped to design and evaluate these tools, better equipped to recognize when they're not working as they were intended, or, in some cases, working as they were intended but not in service of the public good. Education itself is a powerful tool to help ensure that technology is used for public benefit.
Q: In the fall, the Kennedy School returned to in-person learning. Have you taken any lessons from the remote learning period with you back into the classroom?
I was personally really excited to be back in the classroom. For me, the joy of teaching is interacting with students in real time, in person, being able to see people's reactions and engage in dialogue. Especially for the type of work that I do, there's a level of trust and engagement that I think is hard to replicate if you're not physically in the same space.
But I also appreciate the benefits of remote learning, some of which I'm hoping will last as we transition back to in-person instruction. It’s great that students can go back and watch the class recordings if they miss things in lecture and want to hear certain points again. For me it often takes hearing something two or three times to really understand it, so I think having that kind of instant replay has really been a boon for students. I also like the ability online for people to ask and answer questions in chat. It's promoted participation from students who otherwise might not feel comfortable speaking up in class, but are totally comfortable posting questions in chat, which then leads to responses that everyone benefits from. I’m hopeful that we’ll be able to bring some of the benefits of remote learning back to the physical classroom.
—
Image by monsitj; portrait by Martha Stewart