Opinion

The College Dropout Crisis

American higher education has a dropout problem. About one in three students who enroll in college never earn a degree. But a promising solution is staring us in the face: Schools with similar students often have very different graduation rates. This suggests that the problem isn’t the students — it’s the schools.

Here we looked at 368 colleges arranged by what we would expect their graduation rates to be, based on the average for colleges with similar student bodies.

Take Western Kentucky University and the University of Nevada, Las Vegas. At both, the expected six-year graduation rate is close to 60 percent.

But the actual graduation rate at U.N.L.V. is far lower than at Western Kentucky. In human terms, the difference amounts to about 500 more dropouts at U.N.L.V. each year.

That kind of variation is common. If struggling schools could learn some lessons from the high performers, tens of thousands more students could graduate from college each year.

Opinion

The College Dropout Crisis

American higher education has a dropout problem. About one in three students who enroll in college never earn a degree. But a promising solution is staring us in the face: Schools with similar students often have very different graduation rates. This suggests that the problem isn’t the students — it’s the schools.

Here we looked at 368 colleges arranged by what we would expect their graduation rates to be, based on the average for colleges with similar student bodies.

Take Western Kentucky University and the University of Nevada, Las Vegas. At both, the expected six-year graduation rate is close to 60 percent.

But the actual graduation rate at U.N.L.V. is far lower than at Western Kentucky. In human terms, the difference amounts to about 500 more dropouts at U.N.L.V. each year.

That kind of variation is common. If struggling schools could learn some lessons from the high performers, tens of thousands more students could graduate from college each year.

The chart above is based on an analysis done by The New York Times and the Urban Institute’s Center on Education Data and Policy. We undertook this project because the college-dropout crisis is a major contributor to American inequality. Many lower-income and middle-class students excel in high school only to falter in college. They then struggle to get good jobs.

College matters so much because it isn’t just about book learning or the development of tangible skills. It’s one of the first obstacle courses of adult life. The students who complete it typically go on to earn more and live healthier and happier lives, research shows.

For too long, high-school students, parents and guidance counselors have hardly thought about graduation rates when choosing a college. And for a long time on many campuses, administrators and faculty members didn’t even know what their college’s graduation rate was.

But now people are starting to realize the stakes. “The issue of college completion is front and center for presidents and trustees,” said Ted Mitchell, a former Obama administration official who runs the American Council on Education, a consortium of universities.

Some high school administrators have also started focusing on the problem, in part out of frustration. Over the years, they have watched their hard-working, talented graduates struggle in college. “It’s disappointing, to say the least,” said Catherine Suitor, the chief advancement officer at Alliance, a network of charter schools in Los Angeles.

In response, Alliance has put together its own ranking, based on colleges’ graduation rates for students who are black or Latino (as most Alliance students are). It then encourages students to attend a college with a relatively high rate. “A bachelor’s degree,” Suitor said, “is the single most influential determinant in multigenerational change and ending the cycle of poverty.”

To create our expected graduation rates, we looked at the characteristics of a college’s students, including income, race, gender, age and test scores. On average, colleges have lower graduation rates when they enroll more lower-income students, more black and Latino students, more men, more older students and more students with low SAT or ACT scores.

We found that the list of top-performing colleges — those that exceed their expected six-year graduation rate — is diverse in almost every way. It spans colleges with expected graduation rates from 18 percent to 80 percent. It includes private colleges, like the University of La Verne, in Southern California, as well as historically black colleges like North Carolina Central and Fayetteville State. It also includes big public universities like San Diego State, New Hampshire, Virginia Tech and several branches of the State University of New York.

Top-performing schools

Any size
Large schools (1K+ students)
Largest schools (2K+ students)

Note: School size refers to the entering first-year class, not the total student body.

The group of underperforming colleges is also diverse, although all of the outliers in our analysis are public:

Struggling schools

Any size
Large schools (1K+ students)
Largest schools (2K+ students)

Note: School size refers to the entering first-year class, not the total student body.

Giving Students Structure Works

Why do some colleges do so much better than others? Part of the answer involves structure. Students tend to do better when they are following defined academic paths, rather than “aimlessly signing up for classes,” as Pamela Jackson, the provost of Fayetteville State, said. Her university, for instance, no longer allows students to be classified as undeclared, instead assigning everyone a “pre-major” based on the interests they listed on their application.

Many colleges also publish “degree road maps” that guide students through a course of study. And more colleges are pushing students to take enough classes to graduate in four years, instead of thinking of college as open-ended. At the University of California, Riverside, a student-government leader made up sweatshirts with a subtle message for the class that entered in the fall of 2014. The sweatshirts read, “Class of 2018.”

But perhaps the biggest lesson from our reporting is that the colleges with higher rates of student success simply seem to have been trying harder for longer.

They collect data on their students, study that data and use it to remove hurdles for students. They deepen students’ connections to other people on campus, including their classmates (through extracurriculars), professors and advisers. “A lot of it seems like it’s attention to detail in catering to students,” said Leebo Tyler, a recent graduate of Troy, an Alabama university with a higher-than-expected graduation rate. La Verne’s president, Devorah Lieberman, said, “The bottom line is connection — feeling like somebody cares.”

That approach worked for Anyssa Ramirez, who grew up in Dinuba, a small city in California’s Central Valley. Her family could not afford to send her on the 200-mile trip to La Verne’s visiting days for admitted students, so two of her teachers offered to take her and a classmate. The visit persuaded Ramirez to enroll.

The next fall, some of Ramirez’s high school classmates struggled with the transition to college, she said, but she felt connected to her campus from the start. Through La Verne’s day of community service and other programs for new students, she got to know classmates and professors early on. “La Verne holds your hand, but not in a detrimental way,” Ramirez said. By our measure, La Verne is one of the country’s most impressive colleges, with an expected graduation rate of 53 percent and an actual rate of 74 percent.

After Ramirez finished classes in December — graduating in three and a half years with a degree in education — she returned to Dinuba to become a substitute teacher. This summer, she will move to Las Vegas to join Teach for America as an elementary school teacher.

college_004.jpg
Anyssa Ramirez, a recent La Verne graduate, is substitute teaching in her hometown, Dinuba, Calif., and plans to move to Las Vegas to join Teach for America. Max Whittaker for The New York Times

Money Works, Too

Another important factor involves student living. At colleges where more students live on campus, graduation rates tend to be higher. After SUNY Brockport, which is near Rochester, began encouraging sophomores to remain in campus housing, for example, graduation rates rose.

Schools with more students living on campus did better

Average gap between actual and expected graduation rate in percentage points

Schools with more than 50%

of students living on campus

+3.3

40 to 50%

+0.9

30 to 40%

+1.2

-0.1

20 to 30%

-2.1

Less than 20%

Average gap between actual and expected graduation rate in percentage points

Schools with more

than 50% of students

living on campus

+3.3

40 to 50%

+0.9

30 to 40%

+1.2

20 to 30%

-0.1

Less than 20%

-2.1

Note: Averages weighted by enrollment.

For students who live on campus, college isn’t just something they’re doing. It becomes their life. When they have a problem — a bad grade, a bureaucratic frustration, a lack of money — they can more easily turn to their friends, teachers or adviser for help. “Living on campus puts you so much closer to all of the resources that are accessible to you,” said Jessica Burley, who graduated this year from Brockport with a double major in criminal justice and sociology.

As Kim Wilcox, the chancellor at California, Riverside, said: “You become connected in a deeper way. It’s harder to fall through the cracks.”

Building dorms costs money, of course, as do many of the other strategies that help students succeed. So it makes sense that colleges that spend more money on students also tend to outperform their expected graduation rates:

Schools that spent more did better

Average gap between actual and expected graduation rate in percentage points

Schools that spent more than

$4,000 per capita

+2.8

$3,000 to $4,000

+1.7

$2,000 to $3,000

+1.0

-1.1

$1,000 to $2,000

-0.8

Less than $1,000

Average gap between actual and expected graduation rate in percentage points

Schools that

spent more than

$4,000 per capita

+2.8

+1.7

$3,000 to $4,000

+1.0

$2,000 to $3,000

-1.1

$1,000 to $2,000

-0.8

Less than $1,000

Note: Data covers spending on student services. Spending data is for 2011. Averages weighted by enrollment.

Unfortunately, most states have reduced per-student funding over the past decade, and some of them have done so sharply. Several states with particularly low average graduation rates, like Arkansas, Missouri and Nevada, also spend relatively little on higher education.

The most effective way for colleges to spend money may simply be to give it to students — not just to pay for tuition but also to cover living expenses. Research has repeatedly found that financial aid tends to lift graduation rates, said Ben Castleman, an education professor at the University of Virginia. Affordability is “probably the biggest factor” prompting dropouts, he said.

One strategy involves expanding the definition of financial aid to help students with unexpected problems, like a broken-down car or laid-off parent, that can lead them to leave campus and never get back on track. La Verne has set aside $1 million next year for seniors who are close to graduation but encounter a financial hurdle. North Carolina Central has a similar program meant to help students complete their “last mile.”

Recreate the Kitchen Table

But student success is not just about money. Some colleges have figured out how to succeed despite modest budgets. Troy is one example. It spends only about $6,000 annually per student on services and instruction, less than half as much as a typical flagship state university.

The secret, administrators say, is to teach students how to be college students. Virtually every step on a student’s path from application to the final semester presents challenges. And the best colleges often have a plan for each step.

They start with the transition out of high school. Bethel, a university in northwestern Tennessee with a higher than expected graduation rate, requires all first-year students to take a class called the College Orientation Experience.

Hal Fulmer, Troy’s dean of undergraduate studies, offers a sports analogy: If you ask an N.F.L. rookie “what the biggest difference is between college and the N.F.L., that rookie will say, ‘The speed of the game,’” Fulmer said. “The biggest challenge from high school to college is the speed of the game.” In college, students face many more decisions — about what to study, where to eat, when to sleep — than in high school.

To help, La Verne places all freshmen in “learning communities” of about 20 students who take the same two classes in their first semester. The university creates a theme for each community, such as “Teaching in a Bilingual World,” and places students in them based in part on their interests.

20190511_graduationrates_nyt_0144
Students at this year’s graduation ceremony for Fayetteville State University. The school’s graduation rate is significantly higher than expected. Travis Dove for The New York Times

After the initial transition, students can still get lost, and colleges have found the old forms of student advising were often stiff and intimidating. Advisers dealt mostly with academic issues, rather than the full range of students’ problems, and their offices were often closed at night when students had more time to visit them.

At Troy, Fulmer said, the student-support center used to be seen as “the center for students with problems.” He instead wants it to be like a kitchen table in the student’s home. “The kitchen table is where I did my homework, where I got encouragement, where sometimes I got behavior modification,” he said. “It was a safe place.”

The strongest advising programs don’t only tell students what they want to hear; they also remind students of the stakes. Felecia Nave, the provost at North Carolina Central, said that faculty members try to help students “understand what they will face in the labor market, understand what their chances are, if they don’t get a college degree.”

“Educated people are healthier,” said Nancy Zimpher, a former chancellor of the State University of New York. “They raise healthier children. They are better citizens. They make less demand on social services.” For too long, she added, university leaders have been distracted and have been chasing prestige and rankings, rather than getting better at helping students succeed. “That’s our job,” Zimpher said.

Change Is Possible

Six years ago, at her first faculty meeting as the University of Houston’s provost, Paula Myrick Short showed her colleagues the university’s graduation rate along with the rate at several of its regional competitors. As a research university, Houston was more prestigious than some of the others. Yet on the most basic measure of success — whether students were graduating — Houston was behind. “Why are they here, and we there?” Short asked the room, referring to the list of local colleges.

Several years earlier, Renu Khator had arrived as Houston’s president. She came from the University of South Florida, which has a relatively high graduation rate. Khator had made lifting Houston’s rate a “no-excuse priority.”

Over the last decade, Houston’s six-year graduation rate has risen almost 20 percentage points, to more than 65 percent. (The number for Houston in the chart above is somewhat lower, because the federally available data for all colleges is older.) Houston’s graduation rate is still low, given its student body. And some of the increase stems from the fact that Houston now enrolls better-prepared students, with higher average SAT scores. But the university has also made concerted efforts to reduce dropouts.

grad_018
Students at the University of Houston’s graduation ceremony. Scott Dalton for The New York Times

One of the biggest changes, administrators say, has been encouraging students to take a full load of classes so that they can graduate in four years. “Faculty have seen that students can do it,” Short said. Students who commit to doing so become part of a program called UHin4, and the university in turn promises to make space for them in any class they need to graduate on time. A second change involved academic advising. Houston’s advisers have become more proactive about helping students who show early signs of struggling.

Houston happens to be in a region, encompassing Texas, Oklahoma, Arkansas and Louisiana, where high dropout rates are the norm:

How states performed

Average gap between actual and expected graduation rate in percentage points

Maine

Vt.

Mass.

More than 3

+3 to -3

Less than -3

N.H.

Wash.

N.D.

Wis.

N.Y.

R.I.

Mont.

S.D.

Minn.

Mich.

Conn.

Ohio

Ore.

Idaho

Neb.

Iowa

Ill.

Ind.

Pa.

Wyo.

N.J.

Calif.

Nev.

Utah

Colo.

Kan.

Mo.

W. Va.

Va.

Md.

Del.

Ariz.

N.M.

Okla.

Ark.

Tenn.

Ky.

N.C.

S.C.

Ala.

Texas

La.

Miss.

Ga.

Fla.

Hawaii

How states performed

Average gap between actual and expected graduation rate in percentage points

More than 3

+3 to -3

Less than -3

Maine

Vt.

Mass.

N.H.

Wash.

N.D.

Wis.

N.Y.

R.I.

Mont.

S.D.

Minn.

Mich.

Conn.

Ohio

Ore.

Idaho

Wyo.

Neb.

Iowa

Ill.

Ind.

Pa.

N.J.

Calif.

Nev.

Utah

Colo.

Kan.

Mo.

W. Va.

Va.

Md.

Del.

Ariz.

N.M.

Okla.

Ark.

Tenn.

Ky.

N.C.

S.C.

Ala.

Texas

La.

Miss.

Ga.

Hawaii

Fla.

Note: Averages weighted by enrollment. Delaware, Nevada, Utah, Vermont, West Virginia and Wyoming have only one school in the dataset, and Alaska has none.

This pattern may stem in part from factors beyond the colleges’ control — and factors that our analysis didn’t take into account. Perhaps high-school students in this region are disadvantaged in ways that their SAT scores, income, race and other demographic data do not fully capture. But that’s very unlikely to be the entire explanation. It is likely that the universities in the region share some of the same problems, like relatively low funding and a historical inattention to graduation rates.

Houston’s recent progress is encouraging because it shows that poor performance is not inevitable. When college leaders, in any region, decide they’re no longer willing to accept subpar graduation rates, they can usually do something about it.

“There are a number of schools that have been very successful and other schools that haven’t been as successful,” said Peter McPherson, the president of the Association of Public and Land-Grant Universities. “Those schools that haven’t been as successful are saying to themselves, ‘Why can’t I do what those other schools do?’”

How we did the analysis

For each college, we studied a recent class of full-time, first-year students. At about 80 percent of colleges, this class began in either in the fall of 2011 or 2012. At the remainder, it began in 2009 or 2010.

We counted students as graduates so long as they received a bachelor’s degree within six years of starting college, even if they ultimately graduated from a different college. College administrators have long complained, with some justification, that the most commonly cited version of the graduation rate counts transfers as dropouts even if they earned a degree elsewhere. We wanted to eliminate this problem.

Our analysis considered all colleges that reported both this version of the graduation rate and student demographic information and also had at least 300 students in an entering class.

In creating an expected graduation rate, we included only student characteristics, not college characteristics. Regardless of whether a college was public or private, big or small, rich or not-so-rich, we wanted to compare it to other colleges that enroll similar students.

Although we focused on first-year, full-time students — for the sake of clarity and comparability — we recognize that part-time students and transfer students are important parts of the college population and may focus on them in a future analysis.

Sources: Education Department (demographics), Student Achievement Measure (graduation rates)

David Leonhardt is an Opinion columnist. Sahil Chinoy is a graphics editor for Opinion. Erica Blom, Kristin Blagg and Matthew Chingos of the Urban Institute helped conduct the analysis.

Correction: May 23, 2019
An earlier version of a chart with this article misidentified South Dakota State University. It is a public university, not private.