As the economics department changes its curriculum to become more focused on data science, it has updated its mandatory econometrics course, shifting the class’ emphasis from theory to real-world examples.
In the course catalog, the department has replaced “Econometrics and Data Analysis,” with “Introduction to Data Analysis and Econometrics,” this fall. The former course will no longer be offered after the summer of 2018. Starting next year, economics majors at Yale planning to write a senior thesis will now be required to take “Introduction to Data Analysis and Econometrics,” as well as a second semester of econometrics.
“Going back to the 1970s, almost all work in economics was theoretical,” said Ed Vytlacil, the professor of economics who chaired the committee responsible for the course change. “It was hard to get data stats, it was hard to have enough computational power — a very small faction of economics was actually analyzing data. And more and more, the trend is towards working with data and doing applied econometrics.”
According to Dirk Bergemann, chair of the economics department, “Introduction to Data Analysis and Econometrics,” is an introductory course that provides students with the ability to analyze data and explore important economic issues such as income inequality, climate change, health and development through data analysis. The class is like a lab in which students learn to use open-source software on their own computers, he said.
Economics professor Nick Ryan, who is teaching the course this fall, said that one case study the class is covering is a well-known experiment that measures discrimination in the labor market, which is difficult because a number of factors other than an applicant’s identity can result in rejection, such as skills and education. The experimenters submitted identical resumes with different names to employers and found that those with African-American names were less likely to receive callbacks.
In class, Ryan said, students will apply a statistical test to this data to show that indeed there are significant differences in the callback rates by race. That is one example, he noted, in which students learn how a data set was collected and how reliable it is.
“There’s a sort of mental traction that comes with using an example you might be familiar with from your life, and thinking about what is the data, what is the analysis that informs this,” Ryan said. “I think that students can feel that, when in the course they’re analyzing something that’s not just a contrived example but that’s relevant and important to the world.”
Vytlacil said the committee took into account three factors in making the course change. First, he said, whereas major economics departments at other universities require two courses, one in mathematics or statistics and another in econometrics, Yale had previously required only one course that combined both topics in a single semester.
Second, recently there has been a widespread shift in the focus and methodology of econometrics courses, Vytlacil said. Traditionally, a class would spend the entire first semester on theory and then incorporate actual data only at the end of the second semester. But nowadays, students tend to work with data much earlier in the process. Although they might not fully understand the theory at first, they build up their theoretical knowledge as they deal with real empirical questions.
Finally, Vytlacil expressed a desire for students to understand that economics is a much broader field than they may realize. The course will include standard topics such as interest rates and asset prices but also topics not usually thought of as belonging to the field of economics, such as marriage, religion and corruption, he said.
Bergemann agreed, saying the course is relevant not only for economics majors but also for students of the humanities and social sciences. He told the News that a journalist, for example, would need an elementary understanding of statistics and linear regressions to put tables together and understand the impact of certain policy changes.
“Increasingly, I would suggest that being able to analyze data is like being able to speak another language. So logic and programming is basically a language requirement,” Bergemann said. “We really want this to be an intro course open to not only economics majors but beneficial for majors in the humanities as well. We strongly urge them to take it.”
There are currently just under 200 students in the class.
Eui Young Kim | euiyoung.kim@yale.edu