Show Notes
Across universities today, raw GPAs treat an A in advanced econometrics the same as an A in introductory film studies. Standard grading systems inadvertently penalize students who challenge themselves and reward those who course-shop.
In this episode, we speak with Harvard Business School Professor Scott Kominers about his recent paper that looks at how mathematical market design and "eigengrades" can strip out course difficulty to measure actual student performance. The paper proposes a fixed-effects model—using student overlap across courses—to isolate true academic capability. We explore how this simple structural fix could end the strategic game of course-shopping.
More from Not Another Politics Podcast
Why do politicians frequently focus on deeply polarizing issues instead of pursuing broadly beneficial reforms?
The "power of the purse"—Congress’s constitutional authority to control public spending—is often cited as the ultimate check on executive power. But how effectively does legislative funding actually translate into policy influence in modern Washington?
Do Democrats and Republicans really misunderstand each other as much as we think?
While we take a quick summer break, we wanted to re-release an episode that felt timely, given the rise of right-wing populism around the world. More than 25 percent of countries around the world are currently governed by populists, from Recep Tayyip Erdogan in Turkey, to Donald Trump in the United States. Based on these findings, populism is at an all-time high, and taking a significant economic toll, according to a recent paper by Christoph Trebesch and his co-authors.
Since the fall of Roe v. Wade, we've been told a simple story: red states are radically out of step with public opinion. But what if that's not actually the reality? On this episode, we speak with Natalie Hernandez, Yale PhD candidate in Political Science, about her upcoming APSR paper, "Asymmetric Representation: Post-Roe Abortion Policy and Public Opinion in the U.S. States." Using a massive dataset of 155,000 respondents, Hernandez finds policies in Democratic-controlled states are actually furthest from the average voter, and what happens when Democratic lawmakers prioritize wealthy donors and national activists over the median citizen.