Many students attend a college they’re over- or underqualified for, and a new paper published by the National Bureau of Economic Research examines why.

The researchers found substantial undermatch and overmatch, or students at colleges below or above their ability level, respectively. About 28 percent of students in the sample who started at a four-year college probably could have gone to a better institution, and 25 percent of students might have been in over their heads.

While those figures aren’t so different from shares of mismatched students three decades ago, the new report digs into this persistent issue.

Mismatches are driven more by the decisions of students and families than of admissions offices, argue the researchers, Eleanor W. Dillon, an assistant professor of economics at Arizona State University, and Jeffrey A. Smith, a professor of economics at the University of Michigan.

Financial constraints, among other factors, tended to spur undermatching, they found: Students from wealthier families were less likely to have undermatched.

Beyond financial constraints, not being close to a well-matched public institution and more time off between high school and college correlated with undermatching, the study found. Information was also a factor: Having a lot of information about college or role models who had gone to college reduced the probability of undermatching.

Parental education was trickier, the paper found: Students with parents at both ends of the spectrum were less likely to undermatch and more likely to overmatch.

“We assume students are making the best choice of college given the information they have and financial constraints they face,” Ms. Dillon said in an interview. “If students are choosing to be mismatched, that reflects either insufficient information or information that leads them to think that mismatching is a good idea.”

Meanwhile, having information and role models who had attended college increased students’ probability of overmatching, Ms. Dillon said. Students with the most information, she said, were somewhat more likely to have overmatched than to have found a good match.

Implications of Mismatch

Previous research has found that many high-achieving students from low-income families—as well as many black and Hispanic students—do not attend institutions for which they are academically qualified, often opting instead for open-access institutions where they are less likely to graduate.

Other research has determined that carefully timed information can affect enrollment patterns for students at risk of undermatching. The new paper identifies overmatching as one risk of providing a lot of information.

Students are almost five percentage points less likely to undermatch if they live within 50 miles of a public four-year institution that is a “good match,” the paper found.

The timing of students’ enrollment also plays a role. “Starting college more than 12 months after graduating from high school,” the paper says, “raises the undermatch probability by five percentage points.”

Still, it’s important to bear in mind that mismatch is not inherently negative, Ms. Dillon said. Measures of college quality don’t fully capture the reality that students may choose a given college for its athletic teams, religious history, location, or specialty in a certain academic area.

The consequences of mismatch need more attention, Ms. Dillon said. Examining the effects of undermatching and overmatching on students’ graduation rates and future employment is her next task, she said.

For the paper just published, “The Determinants of Mismatch Between Students and Colleges,” the researchers examined a cohort of students from the National Longitudinal Survey of Youth who made their colleges decisions between 1999 and 2002. The researchers gauged students’ ability using the Armed Services Vocational Aptitude Battery test. They gauged the colleges’ quality with two indexes, one incorporating average SAT or ACT scores, rejection rates, faculty-to-student ratio, and faculty salaries; and another accounting for colleges that admit all applicants and do not report SAT or ACT scores.