Executive orders are often seen as one of the U.S. president’s most powerful governing tools — a way to shape policy with the stroke of a pen. vlog’s Lindsay Scheffler ’27 is spending her summer investigating why this picture is incomplete.
Scheffler, a double major in economics and political science with plans to apply to law school this cycle, is spending the summer working with Assistant Professor of Political Science Annie Benn through vlog’s faculty-initiated summer research program. Working alongside fellow researcher Ellie McGoldrick ’27, Scheffler is documenting evidence for Benn’s forthcoming book chapter, which argues that congressional oversight reshapes how executive orders function once they leave the president’s desk.
An executive order directs federal agencies and officials on how to carry out existing laws and policies. “We think of executive orders as just being an authority that presidents have and that they can do whatever they want with them, which isn’t really true,” says Scheffler.
Although presidents can issue executive orders without congressional approval, agencies need some statutory authority to implement them, and they are still accountable to Congress through oversight hearings. Benn’s research asks how that continuing oversight influences the way executive orders are interpreted or carried out in practice.
The project requires Scheffler to review congressional hearing transcripts, documenting how members of Congress talk about executive orders — identifying statements that touch on an order’s legal authority, how agencies are implementing it, or its effects — and assigning each one a code. Those coded transcripts will allow Benn to systematically analyze when and how Congress scrutinizes executive orders across many hearings, providing evidence for the forthcoming chapter.
Scheffler and McGoldrick are also validating artificial intelligence (AI) models Benn is using to help process those hearing transcripts. “These transcripts are really long — it would take a very long time for individuals to sort through all of this information,” Scheffler explains.
She and McGoldrick independently code samples of some of the same transcripts, then compare their results against each other and against Benn’s own coding standard. Once the research assistants and Benn have reached a high agreement threshold, they are able to validate coding done by AI. This hybrid approach leverages AI to review thousands of pages of congressional hearing transcripts while maintaining high-quality output, as validated by human coders.
Unlike work done in a science lab, much of the work for this project happens asynchronously. “We meet with Prof. Benn a couple of times a week, but a lot of my work is pretty independent,” says Scheffler, who is spending the summer living in Hamilton. “I’ve been able to explore new study spots around town that I have never used before.”
Benn assigns weekly goals, checking in with Scheffler and McGoldrick as they work through the transcripts before reconvening to review results.
As Scheffler prepares for law school, Benn’s project provides a look at lesser-known government processes that shape the U.S. legal code.
“I have always found executive orders to be an interesting aspect of our executive branch, and particularly with the current administration, their role is becoming more and more prominent,” Scheffler says. “I had never really thought about how Congress responded to these orders and that the legal aspect of them continues through the other branches of government.”