Why did some colleges rise and others fall in the rankings?
Yesterday, U.S. News came out with their annual rankings of U.S. colleges and universities. While often criticized, the rankings have an impact on college choice, reputation, and even college president salaries. Below is a graphic showing who went up or down in this year’s rankings.
Why did Harvard go down and Stanford up?
U.S. News uses a mix of 17 indicators to rank colleges. Below are some of these indicators for Harvard and Stanford along with the weight assigned to each indicator. Assessing the exact impact of each change is quite complicated. The impact depends on not just the published weight but also on the standard deviation of the factor. It is notable that Harvard slipped by 0.1 points in the reputation survey. The increase in student faculty ratio also seems significant. To go from a student faculty ratio of 5 to 7 implies a 40 percent (!) increase in the number of students (holding the number of faculty constant). According to IPEDS, Harvard’s undergraduate enrollment went up by only about 12 percent between fall 2020 and fall 2021. I suspect the dramatic increase in student faculty ratio is more driven by changes in how faculty or students are being counted rather than actual changes in the situation on the ground.
Why did Colby go down and Vassar up?
Among the top national liberal arts colleges, Colby dropped seven spots while Vassar went up by nine. Looking at some of their statistics, Colby had a modest drop in their graduation rate. Colby also had a twenty percentage point drop in the graduation rate of Pell Grant recipients, while Vassar had a significant increase. This covers classes entering in the fall of 2014 and 2015. Pell Grant graduation rate is multiplied by the portion of students receiving Pell grants before it enters the overall score formula. More importantly, what does it mean when a graduation rate falls by twenty percentage points in one year? Does that mean that something significant happened at the school? Or, is it that the Pell Grant cohorts are so small that one or two students can sway the statistic?
Overall, rank is pretty persistent — particularly at the top and bottom.
The chart below show the share of schools with a ranking in the same decile this year as last year. On average, about 60 percent of schools end up in the same decile (i.e., within 40 spots for national universities and within 20 spots for national liberal arts colleges). There tends to be less movement at the top and bottom of the rankings — in large part because the overall score (which determines a school’s rank) has a bell-shaped distribution with more schools in the middle than around the edges. Therefore, a small change in the overall score can lead to a large jump in rank.
For more details on the the nature of U.S. News ranking methodology and the implications for resource allocation I suggest this article. For prospective students wanting to know how much college matters, I suggest this.