By Richard Vedder
Whiz Kid Matt Denhart examined the four year college graduation rate for 100 public and 46 private research extensive universities, trying to explain the considerable variation in the rate around the mean value of 41 percent (less than one-half the students graduated in four years). (Matt actually did this a month or so ago; he is now teaching English and French to Ghanan orphans as part of his service to the broader community).
The evidence shows little impact from using part-time (as opposed to full-time) faculty, or from having a low student-faculty ratio to encourage one-on-one relationships between students and faculty. Spending per student was a relatively unimportant determinant of graduation success as well. In other words, financial resources devoted to instruction were of secondary importance. But two things DID matter. First, other factors equal, students graduated more in a timely manner from private schools than public ones. Dropout rates are higher in the public colleges. This suggests the cost differential between the two types of institutions is exaggerated (it takes more years to get through a public university on average), and may explain who private schools are flourishing, even gaining market share, despite being relatively costly.
Second, and more important, the higher the average ACT score of the top quartile of students, the larger the proportion of students who graduated in four years. Compare two schools, one with an minimum top quartile ACT score of 28, the other with a minimum of 23. If all other factors are the same, the statistical model predicts the graduation rate would be 24.5 percentage points higher (e.g., 60 vs. 35.5 percent) in the school with the higher admission standards. SAT scores are a relatively powerful predictor of graduation success, which makes the minor boomlet to do away with requiring the scores all the more puzzling and reprehensible.
This is consistent with the feelings of many of us. Harry Stilles, the determined former college professor and state legislator in South Carolina, has been developing analyses showing this sort of result for years. Jackson Toby, retired Rutgers University sociology professor, has completed a book length manuscript arguing that low college standards, in part reflecting non-selective admissions practices, has seriously detracted from not only higher education's performance, but also that of the K-12 system. Given the ease of getting into college, Jackson feels, almost certainly correctly, that younger kids coast through a non-rigorous secondary school curriculum. I like Jackson's work, and research director Andy Gillen, associate director Bryan O'Keefe, and I are trying to help him get it published. And, of course, my colleague at the American Enterprise Institute Charles Murray says students with limited cognitive abilities do not belong in colleges, but rather in vocational schools. Amen.
As the pool of 18 to 22 year old students starts declining in a few years, the temptation will be to lower standards still further, if possible. While some selective schools want to maximize their US News & World Report rankings by turning away kids, an even larger number are revenue maximizers, taking as many students as possible to increase the college's revenue stream and size, and maybe even the income, of both the colleges and the administrators who run them.
By taking intellectually challenged students, colleges play a cruel hoax on many of them, especially ones who borrow large sums for college only to drop out. Enormous resources are wasted. Maybe we need to be tougher in utilizing public funds (or tax-exempt private funds) to fund students whose mediocre academic background suggests they do not belong in college. Maybe we should give everyone a taxpayer subsidized chance at low cost two year colleges, in keeping with the American equalitarian ideal, but be hard nosed as to who can attend the more costly four year institutions.
In any case, hurray for Harry, Jackson and Charles for their heretical but almost certainly correct thoughts.