Data Science Within Harvard

My colleague1 Rafael Irizarry wrote about “the role of academia in data science education” last fall. I agree with most of his comments but his discussion about the “implication for academic programs” failed to highlight the most important strategic questions which Harvard faces. These questions are best considered from the point of view of the key decision-makers involved. Start with my boss’s, boss’s, boss’s boss:

President Larry Bacow: Should Harvard have a separate school for Data Science — perhaps modeled on Harvard Business School — centered around a two year masters program? Perhaps the insert-rich-person-name Harvard Data School? This is a presidential-level decision because starting a new school is something that only happens every few decades.

Dean of the Faculty of Arts and Sciences Claudine Gay: Should FAS have a new department of Data Science? I don’t know nearly enough about the history/politics of FAS to have an informed opinion on this topic. I am certain, however, that a new concentration in data science would be wildly popular among the undergraduates.

Government Department Chair Jennifer Hochschild: Should the Government Department do more with data science? To some extent, this question has already been answered in the affirmative, with the creation of the Data Science Program, but there are many details left to work out, mostly for Professor Hochschild’s successor.

Director of Undergraduate Studies Nara Dillon: Should the Government Department require two methods courses for undergraduate concentrators? Currently, concentrators must2 take either STAT 100 or STAT 104 or GOV 50. Economics, on the other hand, requires two methods courses, generally fulfilled with STAT 104 and EC 1123. My (not unbiased!) view is that, today, political science as a field is just as interested as economics in the use of data to learn about the world.

All these questions are above my pay grade, as we would say in the Marine Corps.


  1. Am I allowed to refer to all 3,000 (?) of my fellow Harvard faculty members as colleagues? I hope so! I have only met Rafa once, when he (generously!) lectured my Gov 1005 class about his work estimating mortality in Puerto Rico after Hurricane Maria.

  2. I don’t know when this requirement was created. Presumably, there was a time in the past during which there were no methods requirements for concentrators. Indeed, other excellent undergraduate programs in political science have no such requirements.

David Kane
Preceptor in Statistical Methods and Mathematics
comments powered by Disqus