I recently ran across this 30 minute podcast by Malcolm Gladwell about how the rankings in US news and world report are calculated. I was surprised to find out that one of the variables (peer opinions), weighs very heavily in the rankings. This should not be included at all!!! This explains so much! I have always wondered why some excellent universities are ranked so low. Now I know!! My conclusion is that US news and world report is simply out to make a buck. They are probably in cahoots with college board. Very disappointing.
This is an interesting topic. I’ll have to listen to the podcast.
I believe the peer ratings are there as “expert opinions” and have to do with a school’s reputation. I’m not sure how I feel about it being such a large part of the rankings (20%), but I do think it has some importance.
Link to how they do their ranking:
https://www.usnews.com/education/best-colleges/articles/how-us-news-calculated-the-rankings
Something I’d point out is that even if you were to exclude peer rankings altogether, the results would look pretty much as they do. Faculty/financial resources, selectivity, peer ranking…they all overlap to some degree.
The bottom line is that institutional assets acquired over generations (endowments, physical plant, faculty, research capacity) tend to draw a disproportionate share of the best HS students (selectivity), often times through favorable need based financial assistance, which requires strong endowments. It’s very circular.
Despite this, I have yet to see someone come up with something better. Revealed preference gets you to roughly the same point. You can look at outcomes. But grad rates are in large part driven by the caliber of admitted student (selectivity). You can introduce most other widely available metrics, weight things differently and arrive at a similar place. You can look at salaries/debt, but at what point? 5 years out? 10 years out? Does salary reflect quality? We need social workers as much as we need attorneys. Arguably more. What about student satisfaction? But can students who have only attended 1-2 schools really compare as well as university presidents via their peer ranking? Washington Monthly came up with their own metrics and wound up with some dubious results. Check out the ranking of National Louis University and its peer schools of similar rank, for example.
The best advice I can offer is that the rankings provide a rough pecking order for schools, but each prospective college student must devise his/her own rankings. What is “better” at #15 may be far worse for student X compared to the school at #80. Use that to your advantage. If more people are chasing schools higher up the selectivity ladder, that offers students some “selectivity bargain” fits further down the list. The rankings are only as important as an individual allows them to be.
2 Likes
It would be interesting to rank the top colleges & universities by endowment and then compare the endowment rankings with the US News Best Colleges rankings. My thought is that the two rankings would be similar.
6 Likes
It was an interesting podcast. I think Ali rankings need to be taken with a grain of salt.
The whole scam against Reed College is fascinating.
Rankings reflect what the reader wants them to be. I’ve said this before— I like whichever ranking places my kids’ schools highest.
7 Likes
I’m not looking for it, but I would be shocked if this has never been posted. 
6 Likes
Would it surprise you to learn that the top 20 national universities on this list are the first 20 private national universities in the US News list?
https://www.reachhighscholars.org/college_endowments.html
No?
HYPSM are top 5.
Edit: since I have the time, limiting only to private national universities, 12 of the 20 would remain within 3 places of where they are in US News rankings.
Biggest upward movers:
ND twelve spots to 7
Rice 7 spots to 9
Emory 6 spots to 14
Dartmouth 5 spots to 8
WashU 5 spots to 11
Downward movers:
Columbia 15 spots to 18
UChicago 10 spots to 16
JHU 10 spots to 19
Weighting endowment/student 60%, standardized test z-score 25% and acceptance rate z-score 15% would probably move every school to within 3 spots of the “rankings”. For the top 20 anyway.
2 Likes
The University of Texas and Texas A&M are ranked numbers 2 and 8 in terms of endowment. Yet, they’re ranked in the 40s and 60s in terms of national universities. That’s insane considering the value they offer. I’m sure a few folks in the Northeast and far West think they’re a bunch of Texas Bubbas when it comes to peer ratings. As the podcast clearly outlined, the peer ratings are bogus. Simply stating an observation.
totally agree
Large publics don’t do as well in endowment per student, which is generally more correlated with USNWR ranking. I have no idea how accurate the list is at Rankings | Largest college endowment per student - Highest endowment per student, but this list indicates the top endowment per student colleges are:
- Princeton
- Soka
- Yale
- Stanford
- Harvard
- MIT
- Caltech
- Pomona
- Swarthmore
- Grinell
- Amherst
- Williams
…
Regardless of whether you use endowment or endowment per student, there is clearly a correlation between endowment measures and USNWR ranking. In earlier posts, I’ve gone in to more depth on this subject. Most of the key rankings components are well correlated with endowment per student, even if they don’t have as direct similarities as categories like “financial resources per student” and “faculty compensation.”
The USNWR weightings are completely arbitrary. I believe the categories and selections primarily relate to increasing earnings for the USNWR company/magazine/website, rather than genuinely trying to create formula to calculate what is the best college. This is particularly evident when you look at how some categories are calculated for areas that one would not expect HYPSM… to do well, such as social mobility, as well as if you look at their history of changes in weightings.
From an individual student’s perspective the rankings are near useless since the weightings used in the USNWR rankings are almost certain to be completely different from what any particular student values in a their college selection. Maybe it can help some students learn about colleges they hadn’t heard about before, but there are many better options than using USNWR rankings for this purpose.
The endowments of Texas and Texas A&M cover multiple campuses (branch campuses) for each school.
1 Like
The methodology is public, and has been for quite some time.
(But I know a good conspiracy theory requires “secret” to get clicks.)
Any rating/ranking system for colleges is arbitrary. Where would one expect a “correct” or “official” rating system to come from?
USNews is just one or many, many ways to evaluate colleges. No one should be using any them as some absolute arbiter to direct their future. USNews has just been successful at marketing. Much like Harvard.
5 Likes
They’re doing pretty well regardless, though. UT Austin has traditionally gotten around half of the funds for the UT system.
3 Likes
Should we divide the total endowment figure for Texas in half in order to estimate an endowment for just UT-Austin ?
1 Like
They definitely are. But if you were to take their net assets/student, there would be about 30 private schools ahead of them. And if you were to factor in acceptance rate and avg SAT at the weightings I listed above they’d be…about mid 40s on a list. About right where they are.
I’m not saying that’s “right” or “wrong” because rankings should be applicant-specific. Just pointing out that there isn’t a lot of sophistication to the US News rankings but they rather crudely make sense. It’s an exercise in quantifying resources and selectivity that roughly sorts things as if money were no object and a hypothetical applicant wasn’t looking for anything in particular.
1 Like
I’m a bit perplexed on how peer ratings as part of a ranking makes it “simply out to make a buck”, or “in cahoots with college board” or “a negative bias”.
Can you explain?
Interesting bit about Reed: in some ways US News is doing them a favor. Reed is a big enough name that people will apply regardless. If they’re ranked number 90 instead of number 15 or whatever among LACs, they avoid getting applicants from students who are chasing a narrow definition of prestige. If all they want are well-qualified applicants who want Reed for Reed, the rankings only serve to eliminate applications from people who are applying for the wrong reasons.
Unless Reed cares about their acceptance rate. And I don’t think they really do.
1 Like
They aren’t. That’s not a legitimate criticism. BUT: with the peer ratings, you do see herding. If a president fills out a score for 100 schools, what will they use as the basis for that score? Last year’s score. It’s better than student surveys, but it’s not great.
But really, we don’t need an annual list of ranked colleges. These are huge institutions/established brands. Nothing really changes that much year to year. Taking an average of the last ten years is probably more accurate than any one year…if you could actually universally rank colleges in the first place.
That’s largely the point of my post. There are plenty of ranking systems with some form of validation. For example, the list at 2023 College Football Rankings - ESPN shows the 2021 pre-season football rankings. The top 5 colleges are as follows. These rankings will be validated during the 2021 football season, as we see how these colleges do in football games.
- Alabama
- Oklahoma
- Clemson (2/3 order differs between AP and coaches)
- Ohio State
- Georgia
If I said the top ranked college for 2021 NCAA football is Stanford instead of Alabama, it would be obvious that I was incorrect, and one could validate this error by measuring performance in the 2021 football season.
However, USNWR and most other college rankings claiming to rank the “best college” do not have any kind of validation of accuracy. If I said the top 5 “best colleges” for 2021-22 were the the list above that is composed of Alabama, Oklahoma, Clemson, Ohio State, and Georgia; nobody could prove that I was wrong because there is not validation of what defines the “best college.” Instead what combination of criteria and weightings define the “best college” varies from student to student and is not something USNWR can define accurately for any particular student, making the list near useless for any particular student.
2 Likes