We already have rampant grade inflation in high schools, and we had it in 1999 when the students in the sample graduated. And we’d rather have kids competing on four years of classwork than one 4-hour test. When the authors introduced a dummy variable for high school (not even really a proxy for school quality), the value of SAT/ ACT in predicting a student’s eventual graduation turned negative.

Just the opposite: This demonstrates the usefulness of the SAT/ACT as measures of high school quality! The average GPA of a high school can be changed at the stroke of a grading policy – Relying on GPA will result in parents pushing for, and getting, grade inflation in high schools.

Yes, rely on GPA for college admissions, but norm the high schools with an objective test like the ACT.

You’re right that we can’t measure “college success” much better than “obtaining a degree.” That’s unfortunate, but you have to remember that almost all previous studies have defined “success” as “first-year college grades” or “one-year retention rate.” Surely you’d agree that success is closer to graduation than it is to those interim measures.

By mocking perseverance–which I tend to think is a pretty important trait for just about everything in life–you’re also shifting the discussion away from college admissions policies to college quality in general. I tend to agree with you on the quality issue, but we still need a way to apportion scarce seats in good universities. This post was about how best to do that.

In theory, that’s sufficient. I’d prefer to talk about reality.

In reality, some degrees are watered-down and border on useless. At some institutions, the majority of programs fall into this category. If we pretend for a second that the degrees they award are little more than certificates of attendance and good standing with the Bursar’s Office, we do higher education reform a disservice.

Bowen, McPherson and Chingos, in a roundabout way, may have just proven not that SAT/ACT scores indicate nothing, but that high school GPA-as-harbinger means higher education is increasingly mimicking the weakness of the average American public high school.