What Sets High-Performing Pell Colleges Apart?

Higher education has long been viewed as an avenue to economic mobility, but lower-income students continue to face disparities in college access and degree completion. Those gaps have persisted even as college enrollment has expanded, raising questions about which institutions are most successful at supporting lower-income students through graduation and into the workforce.

New research from the University of Wisconsin at Madison, published in AERA Open, identifies colleges and universities where Pell Grant recipients enroll and graduate at relatively high rates, face small completion gaps compared to their higher-income peers, and see strong earnings after college.

Drawing primarily on federal data from 1,462 four-year institutions from 2015–16 through 2019–20, researchers evaluated colleges across three dimensions: access for Pell recipients, degree completion and postgraduation earnings. They also examined whether institutions graduated Pell recipients at higher rates than expected based on their student populations, resources and other characteristics.

The researchers identified 149 high-performing Pell institutions that met their criteria. At those colleges, Pell recipients had an average graduation rate of 76.5 percent, compared with a 50 percent median across the broader sample. The average graduation gap between Pell and non-Pell students was 3.8 percentage points, compared with 13 points across the sample.

Taylor Odle, assistant professor at UW Madison and co-author of the study, said the institutions identified as high performing represent a wide range of colleges and universities, challenging assumptions about where strong outcomes for lower-income students are possible.

“Good work is being done across the higher education sector,” Odle said. “What we want to do is show that there is data out there that complicates some of the stories that we’ve always had in our minds, but that data can be used to make very high-stakes decisions in support of these lower-income students.”

Patterns of success: The list includes highly selective private institutions such as Harvard University, the Massachusetts Institute of Technology and Stanford University; public flagships including the University of Texas at Austin, the University of Michigan and the University of Wisconsin at Madison; and liberal arts colleges such as Amherst College, Swarthmore College and Williams College.

The list also includes religiously affiliated institutions and colleges in urban, suburban and rural communities. Nearly all of the high-performing Pell institutions are public or private nonprofit colleges and universities, with three private for-profit institutions also making the list: Chamberlain University, DigiPen Institute of Technology and West Coast University.

“There’s this narrative in the policy space that student success can only be achieved at a certain subset of institutions that are heavily resourced or that are highly selective,” Odle said.

“We have public institutions, private institutions, rural, urban, highly selective and generally much broader access,” he added. “Institutions are showing that they can achieve not only high outcomes but equitable outcomes.”

Despite their strong outcomes for Pell recipients, high-performing Pell institutions did not necessarily enroll especially large populations of lower-income students. About one-quarter of students at the 149 institutions received a Pell Grant, compared with 38 percent across the broader sample.

Odle said higher education policy has historically placed significant emphasis on expanding college access, but increasing enrollment among lower-income students does not necessarily translate into higher completion rates.

“The strategies that we employ to get more students in the door and lower those involvement barriers are not the same strategies we need to use to help them through completion,” Odle said. “There are many institutions that you could look at nationally that enroll a ton of Pell recipients, but their Pell graduation rates are substantially lower, and I don’t think that’s an outcome that anybody would be particularly excited about.

“I’m glad that our policy conversations have moved to the space of ‘It’s not good enough to just get more folks from low-income backgrounds in the door. We need to be sure that they also have an equal chance of completing that degree and are having similar earnings to their non-Pell peers,’” he added.

Researchers also found that traditional measures of institutional quality, resources and student demographics did not fully explain why some colleges performed particularly well for Pell recipients. Factors such as student services spending, financial aid, admissions rates and the racial and socioeconomic makeup of the student body had limited ability to predict which institutions would qualify as high-performing Pell.

“Those predictors actually don’t do a really good job of explaining why an institution appears to be on this list,” Odle said. “There’s something else going on, and we need more political investment, we need more research and we need more private partnerships to find out what that secret sauce is.”

Sky Duke, a doctoral student at Madison and co-author of the report, said the findings suggest researchers may need to look beyond individual programs or interventions to understand how an institution’s broader environment contributes to Pell student success.

“It’s not necessarily going to be just one intervention that’s happening in isolation in that institution, but rather a combination,” Duke said. “That environmental piece is really worth investigating further.”

Beyond access: Duke said the study offers a broader way of thinking about socioeconomic equity in higher education by considering whether colleges provide access to low-income students, support them through degree completion and position them for economic mobility after graduation.

“It’s not necessarily an issue with the students,” she said. “But simply that, given an appropriate environment that can uplift them and support them through completion … these things are possible.”

Duke said identifying the 149 institutions is a starting point for understanding what they are doing differently and whether those approaches could work elsewhere.

“Now that we have the data, [we can] start following up and investigating to see how these things can expand to different institutions that are looking to improve success rates,” she said.

Odle said the findings are particularly relevant in the current political environment, where “we can’t not ask about completion and labor-market outcomes for these colleges.”

“It relates to what we hope people do with this work,” he said. “It’s not just getting low-income students in the door. It’s also graduating them at high rates and making sure that they have positive and relatively strong labor-market earnings when they leave.”

Get more content like this directly to your inbox. Subscribe here.