When American soldiers returned from World War II, the new GI Bill suddenly opened the doors to higher education for a generation of men whose families could never have afforded tuition. 

Free tuition should have meant that men from all walks of life would attend every type of institution, from the most affordable state college to the most expensive Ivy League university. It didn’t. 

For the past five years, economist Santiago Pérez has studied the socioeconomic makeup of students at elite institutions. A new working paper shows that neither free tuition nor the introduction of standardized testing had any impact on the backgrounds of students attending elite institutions for the last hundred years.

“We find that in these elite colleges, students tend to be overwhelmingly from relatively rich backgrounds,” said Pérez, an associate professor of economics in the College of Letters and Science at UC Davis. 

A century of data from elite institutions 

Since before the COVID-19 pandemic, Pérez has been studying elite higher education with co-authors Ran Abramitzky and Jennifer Kowalski at Stanford University and Joseph Price at Brigham Young University. They have been exploring two key policies that logically should have increased the share of students from low-income families in elite colleges and universities. 

The idea started with debates over standardized testing. Some argued that entrance exams like the SAT create an opportunity for talented students from disadvantaged backgrounds to stand out. Others argued that students from low-income backgrounds could never afford the test-prep and other advantages enjoyed by students in rich families to get higher scores.

There hasn’t always been a SAT, which made it possible to see if there was a difference in the student makeup of elite institutions before and after it was first given in 1926. What Pérez and his colleagues needed was data.

Over the course of three years, they painstakingly collected publicly available data from 65 elite colleges and universities across the U.S. spanning roughly the last 100 years. The data, which included students’ family background and other demographics, are partly based on a large-scale digitization-done-by-hand of historical student records for nearly 2.5 million students.

The elite colleges and universities in the data include Ivy League schools like Harvard and Princeton and “Little Ivy” liberal arts colleges like Bowdoin and Bates, as well as Ivy League equivalents for women like Radcliffe. Elite public universities, like UC Davis and other University of California campuses, are also included.

From the SAT to the GI Bill

The analysis showed that over the last century, the share of students from low-income backgrounds has largely not changed.

“If you look at what percent of the class in elite private colleges came from lower-income families instead of higher income families after the introduction of the SAT, we find that there is basically no change,” said Pérez. “There's nothing going on there.”

This result seemed to settle part of the question of whether identifying talented students from any background increases access to an elite education. But what if the problem isn’t a lack of talent but a lack of money? 

Pérez said that intuitively, the GI Bill should have led to more men from poor backgrounds in all institutions, even the elite ones. This wasn’t true at all. Getting into an elite institution had nothing to do with being able to afford it. Instead, it had to do with academic preparation. Returning soldiers from wealthier families who had received the best academic preparation were also in the best position to take advantage of the new benefit.

“For someone who had only seven to eight years of schooling, getting into any college is going to be much, much harder,” said Pérez. “This example also allows you to learn more generally about the limited role that financial barriers might play in why we see an underrepresentation of lower-income students in these elite colleges.”

Why an elite education stays elite

While the analysis shows that neither the SAT nor the GI Bill increased socioeconomic diversity in elite institutions, it can’t explain exactly why. Pérez and his colleagues can, however, rule out some possibilities. 

One might be that there simply are not enough high-achieving low-income students. The problem with that explanation is that if you look across the 20th century, elite institutions did not put a great emphasis on academic ability. A student’s family mattered far more.

Pérez and his colleagues have expanded this line of research to studying legacy admissions, which are students whose parents also attended the institution. One early finding is that a student with a father who went to Harvard is 400 times more likely than all other students to also go to Harvard. 

The analysis found one exception to the lack of socioeconomic diversity, and that is the University of California. UC campuses were unique among these elite institutions because over the last 100 years the socioeconomic backgrounds of their students have become more diverse. 

“If you look at, let's say, Berkeley in the 1920s, it actually looks much closer to Harvard in terms of socioeconomic class of students than it does today,” said Pérez. “The UC system is actually a real success story in the way that these proportions have become much more balanced relative to what they what they used to be.”

In a separate working paper, Pérez and his colleagues have also found that socio-economic background shapes faculty in higher education. Analyzing the largest existing dataset of U.S. academics’ backgrounds and research output, which they also painstakingly created, they found that professors from poorer backgrounds have been severely underrepresented for seven decades across higher education. This is especially true in both the humanities as a field of study and at elite universities in general. 

Pérez said that socioeconomic diversity matters for many reasons, but discussions on the subject usually focus on equity and efficiency. Leaving people out creates a lack of equity as well as a lack of efficiency because we miss potential innovators, politicians and professors who could improve society for everyone.

“Background really matters for how people think about the world,” said Pérez. “It's going to really matter for what topics are studied and which laws are passed.”

YOU MAY ALSO LIKE THESE STORIES

Forget What You Know about Immigrants and Crime

New research finds that for the past 150 years, immigrants in the United States have had lower incarceration rates than the U.S.-born, and since the 1960s that gap has significantly grown.

Weaving Facts Through the Complexity of Global Migration

Migration has also long been a flashpoint for anger and prejudice in every country and context, and these responses can blur the reasons why people relocate their lives and what their arrival means for the places where they settle. The UC Davis Global Migration Center is a research collaboration that weaves facts and human experience into the complexity of how we think about migration, both in the U.S. and globally.

Stories Archive