Sixty-three years separate the youngest member of Congress, Rep. Maxwell Alejandro Frost of Florida (29 years old), from the oldest, Sen. Chuck Grassley of Iowa (92 years old). When Grassley was a freshman congressman in 1975, his oldest colleague was Sen. George Aiken — born in 1892. A single Congress, let alone several, is the story of a multitude of overlapping lifetimes. There’s a correspondingly large and important story to be told about the aging congressional agenda. The twists and turns in that story — new trends emerging, tidal shifts in political fashion — leave footprints in data. If one pays close enough attention, those footprints can lead to a hidden wealth of policy stories and understandings.
Among the students helping tackle the challenge is UT government and history junior Ava Snoozy, who helps lead a team of undergraduates tasked with categorizing thousands of congressional bills. She works under the mentorship of UT Austin professor of government Sean Theriault through COLA’s Undergraduate Research Apprenticeship Program (URAP).
The backbone of the work is the Policy Agenda Project (PAP), started in 1993 by political scientists Frank Baumgartner and Bryan Jones and now led by Theriault. PAP provides a trove of datasets that tell the story of Congress and how its policies age and change. By looking closely at the amount of time legislators spend considering policy recommendations and other measures of congressional attention, PAP researchers uncover the day-to-day priorities that shape Congress and the nation.
Over the years PAP has developed datasets tracking roll call votes, Supreme Court opinions, and even the phrases that presidents utter in their annual State of the Union addresses. But the usefulness of each of these datasets is dependent on the accuracy and organization of the data they contain. That’s where URAP comes in.
“The congressional bills dataset was never particularly good,” says Theriault. “And so just over the last year and a half or two years, we’ve been really trying to make it a top-notch source of data.”
Snoozy plays a lead role in the congressional bills project. She runs a team of undergrad students in sorting thousands of bills into categories that can then be further analyzed computationally.
“We read over the bills and code them in one of 21 categories, which are broad,” says Snoozy. “Those bills are then further categorized into subcategories, which just makes sorting and looking through the catalog of bills a lot easier for users of the dataset.”
The major categories include large buckets like immigration, transportation, governmental operations, and defense, says Snoozy. Within a large bucket, such as defense, there can be smaller categories, like military procurement and weapon systems.
At the start of the semester, Snoozy’s team was given a tranche of bills, perhaps 10% of the total, to numerically code by topic. This initial step teaches the computer software they use how to handle the remaining 90%.
Theriault stresses that categorizing bills is an iterative process, one that requires a lot of back and forth and judgement calls. He uses the example of a grant program that tries to get states to install electric-vehicle charging along highways. Is that a transportation issue? Or a federalism issue? Perhaps it best fits under environment or energy?
“Talking through those cases is where the real learning happens,” says Theriault. As students gain confidence with their coding, both the rate and the degree to which coders agree with one another increases. The pace of the work increases as well.
The student work is an essential element of the overall project goal of deepening our understanding of how congressional priorities have shifted over time. “If you were to look at the differences in military spending — what changed from the 1970s to the 2000s, for example — you get a very good picture,” said Snoozy. “And you can see the differences in these categories over time in a very clean and organized way.”
One of the challenges of the project is that many older congressional records are still being recoded and standardized, and so there are some trends that simply can’t be observed yet. That makes the students’ commitment to consistency and standardization especially essential, said Theriault, since they’ll likely end up handing the project on to future cohorts of students. Already, though, the dataset allows scholars to compare congressional priorities across decades with confidence. It’s a model of intellectual rigor that stands out in a time where trustworthy information about politics is at a premium.
“For anyone trying to be engaged in political life and trying to understand what our current situation is, the fact that we have this long-term data set that focuses so strongly on maintaining the integrity of our methodology is really important in a world where sometimes facts and figures are construed in manipulative ways,” said Snoozy.
Theriault has seen the real-world impact of his work on students who have gone on to careers as state representatives or lawyers having sharpened their analytic skills under his tutelage.
“One of the students that worked in my lab is now a member of Congress himself,” said Theriault.
For Snoozy and her fellow undergrads, URAP is an intellectual gateway. It’s an experience in applied knowledge that also shows how collaboration and respectful debate are core to the research enterprise. The very work of categorization involves a willingness to advocate for a position over proffered alternatives.
“No one’s yelling at each other,” says Snoozy about the vibe on her team. “But sometimes it can get pretty intense. I think just being able to stand by your decision and explain articulately why you came to that conclusion based on the information you’ve been able to glean, and then either convince other people or adjust your reasoning, is a practical skill that sometimes undergraduates don’t have the chance to practice.”
“Being able to say what you think and why you think that way—and being willing to change your opinion when presented with better evidence—is a skill that applies far beyond political science,” Snoozy says.