The Princeton University Chapel in 2015. The university was the No. 1 national university for the eighth straight year on U.S. News & World Report's annual list. (Yana Paskova for The Washington Post)

This time of year always brings a robust debate over college rankings — which ones matter most, and whether to trust what they value. Too often, rankings have asked America’s college and university leaders to face the tough choice of investing in students from low-income families or weeding them out in a quest for prestige.

These decisions are critical for both our nation’s economic future and the futures of low-income students and their families. Jobs that once required a high school diploma now demand a college degree. But only 14 percent of students from low-income families graduate with a bachelor’s degree. With an expected shortage of roughly 11 million college graduates by 2025, we cannot remain economically competitive if we focus on middle-class students alone. Colleges ought to pull out all the stops to help more low-income students graduate. Historically, the incentives in place have not rewarded colleges for doing this.

The now well-documented story of Georgia State University helps to illustrate the point and may signal a welcome shift among rankings. During the past decade, Georgia State made a major effort to help low-income and minority students graduate with a bachelor’s degree. They didn’t just cherry-pick the best low-income students. Rather, they improved their advising and other support for students who had the potential to graduate, but who all too often dropped out. As a result, they eliminated income and racial gaps in graduation rates while raising their overall graduation rate by 23 percentage points. This is practically unheard of in higher education.

What was the result? In 2014 the school dropped more than 20 places in the next edition of the U.S. News & World Report annual rankings. Of course, U.S. News didn’t set out to ding them. But in an era where colleges are often valued more for the number of students that they turn away, Georgia State’s decision to admit more students with lower SAT scores outweighed the positives of helping these students graduate. Exclusivity, all too often, becomes a proxy for quality, because quality is very difficult to measure.

Our system has been predicated on the assumption that students don’t graduate because they couldn’t cut it, not because the university may have let them down. Students, of course, can’t be let off the hook — they need to work hard, get to class, and not let the distractions of their first taste of freedom prevent them from buckling down at exam time. But placing the blame entirely on students overlooks opportunities to fulfill higher education’s economic promise and mission.

Fortunately, we’re beginning to see early indications of a shift. In U.S. News’ most recent rankings, which considers social mobility and outcomes, and diminishes the value placed on selectivity, Georgia State jumped from 223 to 187. The institution was rewarded for serving low-income students well. And while this is not a sea change, it is a sign of progress.

Because improving outcomes for low-income and minority students has to start with transparency. Today, results for low-income students are largely hidden. Pell Grant awards are typically used as a stand-in for low-income status, but the number and graduation rates of Pell students at various universities are not readily available. For example, I work with a group of 11 universities across the country that in the past four years have increased the number of low-income students they graduate by 29 percent. That equates to 7,000 additional Pell graduates per year. These universities are pretty sure they’re doing a good job, but because most institutions don’t make their Pell graduation numbers public, it’s impossible to know for sure how they stack up.

As a nation, we need to know which colleges are serving low-income students well. The goal should not be to praise or shame, but to find out what’s working and then replicate it. Here are three ways we can start measuring what we truly care about:

1. Publicize measures of low-income student progress to identify those schools that graduate large numbers and percentages of low-income students. Colleges know this information already, but do not share it publicly.

2. Develop rankings that highlight and reward the kind of behavior our country needs more of. Rankings might, for example, showcase universities that admit and graduate (in four and six years) the highest number and highest percentage of low-income students. Or they might shine a light on universities with the smallest gap in performance between low-income students and the rest of the student population. There is room for optimism here, as ranking organizations increasingly take into account graduation rates, Pell rates and student satisfaction.

3. Make sure that awards and accolades for serving low-income students reward scale, not exclusivity. Yes, we want elite institutions and small liberal arts colleges to serve low-income students, and we should applaud their efforts to do so. But they serve relatively small numbers of students. We have to find ways to deliver high-quality degrees at the scale our country needs, which means more emphasis on large universities and less on the Ivies.

Transparency doesn’t require a massive policy shift. State and federal policymakers can start by asking university officials to share the retention and graduation gaps between their low-income and high-income students. This would push more data into the public sphere — where it would be accessible to parents, policymakers, researchers, and journalists — and encourage universities to close those gaps.

Data sharing would jump-start the process of unmasking differences in college success rates between low-income and high-income students. And it would begin to de-risk the efforts of institutions that are working to make a difference for low-income students. If we’re going produce the college graduates America needs and give all students a shot at a better economic future, it’s time to measure what matters.

Bridget Burns is the executive director of the University Innovation Alliance, a national consortium of 11 large public research universities collaborating to improve outcomes for students across the socioeconomic spectrum through innovation, scale and diffusion of best practices.