I don’t think this is really a sensible claim to make. Obviously if they’re comparing the 4.0 students to the 3.0 students…more of the 4.0 students are likely to go to college. Many of these 4.0 students continue to do well in college, and graduate from college. Some of them don’t. The ones that graduate fall into the income bracket known as “college graduate” while the ones that didn’t graduate end up falling into the bracket known as “high school graduate.” Clearly, a college degree tends to increase lifetime earnings.
Now if we’re comparing students that graduated from high school, -but- never graduated from college, we’d have a more sensible comparison. I highly doubt that there would be any kind of significantly discernible distinction in the lifetime earnings of students that graduate from high school with a 4.0 versus the ones that graduated high school with a 3.0.
What about the 4.0 students that DON’T go to college, and the many 2.0-3.0 students that DO go to college? Now we’re saying that the 4.0 students are going to earn both less and more than the students with lower high school GPA. I think this is a prime example of defining the parameters in such a way that the findings fall in line with a desired conclusion. I really don’t think that anything truly meaningful can be drawn from this.