The Graduation Rate Gap: Minding What Matters?

One of the dirtiest secrets in higher education is that getting in--the focus of 18 years of hard work in some families--is no guarantee of college success. The sad truth is that at many public four-year colleges and universities fewer than half of entering freshmen complete a degree within six years. And in some cases those low average rates conceal incredible disparities based on race and family income.

Kevin Carey of Education Sector has done a remarkable job at turning the nation's attention to this often overlooked problem in recent years. He helped create College Results Online which enables the average Joe to spend hours of fun picking out individual schools and seeing just how low their graduation rates can go. (Seriously, go check it out-- it's FUN!)

Most recently Carey issued Graduation Rate Watch, a report that uses six years of federal data to look at changes over time in the black-white gap in college completion. His primary point: between 2002 and 2006 some schools made significant strides in closing their gaps, while others did not. And in fact, some schools had (and have) no gap at all. Therefore, Carey concludes, the answers are clear: too many colleges are indifferent to the plight of minority students, they know what they need to do, and we just need to push them to do it.

Kevin and I agree on most things. We agree that the feds and states are paying far too little attention to low rates of college success. We agree that disparities in degree completion rates are unacceptable. And we agree that institutions of higher education need to be a part of the solution But on one crucial point we disagree: what should be done.

Drawing on a case study of a single college (Florida State) and its lovely college-support program CARE, Kevin contends that the solution to closing the college achievement gap lies in "creating incentives for institutional leaders to act on the knowledge that already exists." His goal in making this kind of statement is to get educators to step up-- they know what needs to be done, and what remains is to take action.

But the researcher in me hesitates to endorse this plan. Here are a few reasons why:

1. There is little evidence that differences in students' college graduation rates are primarily or even substantially attributable to school-level factors. Go back to the Coleman report--his intelligent hypothesis was that inequities in schools due to segregation were driving black/white differences in educational outcomes. But instead he found that family background and other student-level characteristics drove far more of the variation in student achievement. Yes, some school-level factors matter--among them teacher quality--but empirical evidence indicated that changing schools themselves would not diminish educational inequality much.

Why would we think things would be different in higher education? The typical argument goes like this: We know that institutional choices and policies (financial aid, academics, advising, faculty, incentives, student engagement, etc) matter in terms of student learning, retention, and attainment. Given our diverse, decentralized higher education "system" there is substantial inter-institutional variation in the quality of those choices and policies. Therefore, "institutional effects" should be significant and real.

Well, they may be significant, they may be "real," but there is little hard data to say that they are large and should be the main focus of policy reform. I'm not saying that colleges and universities aren't different from one another in ways that matter, but because higher education is not compulsory and its gates are closely guarded by the gods of test scores and $$ the level of stratification in the college entry process is so intense that students attending different schools are very different from one another. There have been very few studies of institutional effects in higher ed that deal with that selection issue appropriately. I strongly suspect that if a good study were done (and given data constraints it'd be hard to do), we'd find that policymakers would get the biggest bang for their buck by working to reduce race, class and gender disparities in who goes to college and where they go.

2. Another important concern: There can be unintended consequences when you push colleges to close achievement gaps. Take this scenario: You've got 30 percentage point gap in college completion, with 40% of black students finishing a degree in 6 years, compared to 70% of whites. You want to close that gap. So you tell your admissions director to solve the problem. And he does: by being more selective in which black kids he admits. Admitting only very highly-prepared black students may mean a decline in the number of black kids on campus, but it's likely to increase the completion rate of those who are admitted. And if he coupled his efforts with a greater effort at targeted recruiting of minorities (including directing merit-based scholarships at them) , he might succeed in simultaneously increasing diversity and completion rates. Not a bad thing per se (perhaps) but definitely not what Carey has in mind...

3. There is not a shred of evidence that FSU's CARE program actually works to increase student retention or achievement. No evaluation that takes into account who they serve to start with, nada. Yes, they've got data indicating that CARE students start with lower entering test scores than other students and end up with higher retention rates, but NO--this doesn't automatically mean that success is due to the program. What if CARE students are simply more motivated than other students-- meaning that they were more likely to be retained from the start? We don't have any way to account for that in the absence of a good ol' random assignment trial where we assign students at random to participate in CARE. Now that would be worth doing!


Of course I understand that Kevin's goals are different from mine. Ultimately he operates in a world where being 51% sure that you're right when making policy recommendations is sufficient-- after all, if we don't advocate for changes in colleges, there might not be attention paid to higher ed at all (or dollars spent). But in my little research world, I need to be much more confident that the recommendation I'm making will be effective in creating the kind of change we aim for--AND, I want to focus on what will make the biggest difference. And there's simply little hard evidence to suggest that adopting a program like FSU's would help to close the black/white gaps in completion at other schools.

Why not--instead of "tinkering at the margins" with school reform--tackle the larger and more systemic problems like ending poverty and racial segregation? (Why is it that educators and educational policymakers have a hard time seeing what happens outside of schools as being part of their domain?) Why not make higher education compulsory, and in doing so greatly increase the motivation to make high school meaningful? Why not re-envision the entire system and create a p-16 structure which compels higher ed and secondary ed to take an interest in one another?

The last thing we need to do is fan the flames of fears about accountability, encouraging colleges and universities to turn even more inward, engaging in ego-think, and directing their attention at self-preservation rather than student success. We do want them to do more-- and they can be an important part of the process of change-- but they need direct their attention at the things that matter most.