The Education Trust has issued a study showing that Pell grant recipients (low-income students who can receive as much as $5,775 per year from the federal government) graduate at a lower rate than non-Pell grant recipients. That fact is no surprise, but the distinctive aspect of this study is that it puts some blame on the institutions.

The report argues that the gap found between Pell grantees and other students in graduation rates (14 percentage points) reveals only part of the problem. When one looks at individual schools, the difference in graduation rates between Pell grantees and non-Pell grantees is only 5.6 percentage points. That stems from the fact that different schools have different graduation rates. So while in most schools Pell grantees graduate at a lower rate than non-Pells, in some schools both rates are low.

The Education Trust concludes, first, that ”institutions with considerable gaps in completion must do more to improve their outcomes for Pell Grant recipients.”

Second, colleges with low graduation rates should increase them.

I don’t necessarily disagree with either of these conclusions except to ask 1) How much investment should be made in the education of poorly-performing students? and 2) Will the goal of raising graduation rates further weaken academic standards?

I have another question about the study, however. A 2012 study by the Pope Center found that nearly 60 percent of all students receive Pell grants—and one must assume that not all those students are needy. Furthermore, the grants range widely in size. Students can get a $500 “leg up,” the full $5,775, or something in between. Recipients of the smaller amounts may be quite different in background, motivation, and current circumstances from those who receive the most.

The Education Trust study does not differentiate among grantees. It would be interesting to compare the college careers of those who receive the full Pell grants with those of other students. That may show the problem of low graduation rates to be more intractable than this study implies.