Here’s a new college ranking, based entirely on other college rankings

College Search & Selection
81

@Alexandre Berkeley is definitely stronger than Duke and many other college top 10 schools when looked at as a research university. It usually ranks just below Harvard, Stanford and MIT in most international rankings that measure research output, influence, departmental strength etc.
However when it comes to college rankings things are different and Berkeley is not really top 10 college. I agree that Duke is a top 10 college.

@IsaacTheFuture Duke’s endowment per student figure was $491,000 in 2015 not $715,000. For 2016 the figure is around $460,000 since the endowment lost $500million.

82

@Penn95 Duke also receives $3.4 billion in the Duke endowment, which is not included in it’s regular endowment. That boosts it’s endowment by about $250k per student.

83

Penn95, those rankings you refer to are generally regarded as unreliable by the experts. Besides, there are many rankings out there, and several of them rank Cal above Duke. Those in the know would find it silly to claim that Duke is superior to Cal.

IsaacTheFuture, only one third of the Duke Endowment is is allotted to Duke University.

84

@Alexandre You’re correct, my mistake.

That would mean their endowment, when including the Duke endowment, would instead be $560,870.

85

That is correct IsaacTheFuture, and you should not forget to include state funding for Michigan and Cal when comparing endowment to that of Duke. Michigan is comparable to a private university with an endowment of $400,000/student and Cal with the endowment of $300,000/student. Cal has no medical school, so it’s endowment is more evenly distributed than that of Duke’s or Michigan’s, which support huge medical complexes. In other words, private elites (like Columbia, Cornell, Duke and Penn are not wealthier than public elites (like Cal, Michigan and UVa).

Also, Duke’s 6:1 faculty ratio does not include graduate students. If you include graduate students, the student to faculty ratio would increase substantially. For example, Duke’s College of Arts and Sciences (Trinity) enrolls 6,500 students (undergraduate and graduate) and has a faculty of roughly 600 professors. That’s an 11:1 ratio. The Duke College of Engineering enrolls 2,200 students (undergraduate and graduate) and has a faculty of 130 professors. That’s a 17:1 student to faculty ratio. Those ratios aren’t much better than Michigan’s. Michigan’s College of Arts and Sciences (LSA) enrolls 20,000 students (undergraduate and graduate) and has a faculty of roughly 1,400 professors. That’s a 14:1 ratio. The Michigan College of Engineering enrolls 10,000 students (undergraduate and graduate) and has a faculty of 480. That’s a 20:1 student to faculty ratio. Like I said, and have been saying for years now, the difference in student to faculty ratio at elite private universities and elite public universities is negligible at best. The main difference is that private universities do not include graduate students in their ratio. Sadly, many posters on CC know that fact and still insist on propagating the lie.

The suggestion that private elites are superior to public elites is not substantiated by the facts, and certainly not one that is supported by academe or industry.

86

@IsaacTheFuture ah you are right hadn’t thought of the Duke endowment, a portion of which is allowed to Duke.

@Alexandre Michigan has an endowment per student around 230k and UVa around 260k, not 400k, right? I wouldn’t go as far as to say that Michigan or UVa are as wealthy as Columbia, Penn, Duke, but they definitely have plenty of resources to provide the students with a great education.

87

Rightly so. The Common Data Set, Section I (“Instructional Faculty and Class Size”), item I2, instructs colleges to “exclude both faculty and students in stand-alone graduate or professional programs” when reporting S:F ratios. Duke’s 2015-16 CDS apparently follows this instruction. Michigan’s CDS apparently does not. Michigan does adjust its faculty number downward to account for graduate-only faculty, reporting a total of 2,512 FTE undergraduate instructional faculty. However, it also reports 37,032 students (which appears to be much higher than the number of FTE undergrads). By my reckoning, if Michigan were following the CDS instructions correctly, they’d be reporting 27,609 FTE undergraduate students. I base this number on the full-time/part-time student counts reported in section B. This would result in a Michigan S:F ratio of 11:1, which would be better than the reported 15:1 but still not as good as Duke’s reported 6:1.

As for institutional wealth comparisons, as I see it, it might make sense to compare endowments alone, or to compare total revenues alone. I don’t see how it is appropriate to compare endowments adjusted upwards by revenues from only 1 source (state appropriations). Isn’t that like saying Jones has a net worth of $700K and Smith has a net worth of $200K, but Smith is as rich as Jones because Smith also gets a social security check that is worth as much as Jones’ extra $500K of investments? Well, ok, but if you’re going to add Smith’s SS check then let’s also consider Jones’ trust fund, etc. etc.

USNWR, by the way, doesn’t consider either endowment or revenue in its ranking. It does consider education-related expenditures as 10% of the total ranking.

88

Pretty accurate list. But Florida above Texas and USC?? maybe for a wet t-shirt contest but not academically.

89

“Michigan has an endowment per student around 230k and UVa around 260k, not 400k, right?”

That is correct. But you should also factor in state funding. Michigan receives $300 million/year and UVa $150 million/year in state funding. Universities use only 5% of their endowment annually. So the $300 million Michigan receives in state funding annuallt is the equivalent of an additional $6 billion in endowment and the $150 million UVa receives annually is the equivalent of an additional $3 billion in endowment. Those should be added to their existing endowments of $10 billion and $7.5 billion.

“I wouldn’t go as far as to say that Michigan or UVa are as wealthy as Columbia, Penn, Duke, but they definitely have plenty of resources to provide the students with a great education.”

When you include state funding, Michigan and UVa are as wealthy as Columbia, Cornell, Duke and Penn.

90

“As for institutional wealth comparisons, as I see it, it might make sense to compare endowments alone, or to compare total revenues alone. I don’t see how it is appropriate to compare endowments adjusted upwards by revenues from only 1 source (state appropriations). Isn’t that like saying Jones has a net worth of $700K and Smith has a net worth of $200K, but Smith is as rich as Jones because Smith also gets a social security check that is worth as much as Jones’ extra $500K of investments? Well, ok, but if you’re going to add Smith’s SS check then let’s also consider Jones’ trust fund, etc. etc.”

I don’t see how that is relevant tk. All universities, public or private, have similar revenue streams. State funding is a unique source of revenue, exclusive to public universities . Can you name a source of revenue that is exclusive to private universities?

“Rightly so. The Common Data Set, Section I (“Instructional Faculty and Class Size”), item I2, instructs colleges to “exclude both faculty and students in stand-alone graduate or professional programs” when reporting S:F ratios. Duke’s 2015-16 CDS apparently follows this instruction. Michigan’s CDS apparently does not.”

Not really tk. By definition, universities should exclude faculty and students from stand-alone graduate or professional programs, such as medicine, law, business (at universities that do not offer business degrees to undergrads, like the Harvard or Yale or Chicago etc…). However, faculty and graduate students enrolled in programs that have undergraduate degree students should be included. Think about it. Don’t graduate students exhaust the resources of a department and faculty? If so, should they not be included in the ratio? Michigan calculates the ratio properly, as requested by the USNWR, Duke does not. Private universities (not just Duke) deliberately leave out thousands of graduate students. Michigan’s 15:1 ratio accurate. Duke’s 6:1 is not. Most elite private universities have student to faculty ratios in the 10:1-15:1 range.

91

@Alexandre So i did a little digging, and found out that UMich received about $295 million in state funding.

http://vpcomm.umich.edu/budget/fundingsnapshot/5.html

That would only increase its per student funds to $236,035, impressive, but hardly what some of the top privates are getting. The UCs receive about $2.374 billion, and on a per person basis, that’s only about an extra $9,432 (bringing it to a total of $115,311. UVA receives about $6,512 per student in federla funding, bring their total up to $349,018 so I don’t know where you’re getting that $400,000 figure from.
Now this does not include the money from the budget or federal funds for R&D research (which I already provided independently), but if we are going to include that into Michigan’s & UC Berkeley’s tally, then we have to do the same for Duke and other schools, which will only increase the tally. By including revenue outside of endowment in their tally without doing the same to other schools, you are being willfully biased.

As for Berkeley’s lack of medical school, you do realize that that benefits them, right? Having no medical school means the resources can be distributed to a smaller pool of people, which in turn means everyone gets more. Now of course, if Berkeley did have a medical school, they would definitely have more monetary donations as a result of simply having more people graduating due to the extra school; however, I highly doubt these added donations would catapult it’s numbers to Duke’s, UPenn, or even UMich’s levels. It definitely doesn’t make up for the hundreds of thousand dollar discrepancy. BTW, Duke’s medical schools is relatively small (453) compared to other school’s and compared to it’s student body at large (only 3% of it’s student body). Even if we assume that Duke’s medical program is sucking up half of the endowment alone - which is insane circumstance and definitely not the case - that still puts it’s endowment at about $290k, still above Berkeley. The notion that Duke’s medical department accounts for it having an endowment nearly 5 TIMES the size of Berkeley’s, on a per-student basis, is just preposterous.

As for the academic staff, tk21769 already handles that. Yes, Michigan’s numbers would improve, but still short of Duke’s and many of the other top Ivies. In fact, Duke calculate it’s faculty number’s incorrectly. They have 1120 full time instructional faculty (excluding row J of I1), plus another 61 part time faculty (excluding row J of I1); divide that by 3 (because they say count 1/3 of part time), and that puts their instructional faculty at 1140. That would still round up to 6 for a S:F of 6:1, however.

I’m not putting down UMich, UCBerkeley (both of which are two of my top schools), or any other public school down in any way. But no statistical evidence suggests that either are superior to Duke, UPenn, Columbia, Cornell, or any other similar elite private university. There’s just too much of a gap in financing and smaller class sizes (though they do stack up when it comes to student feedback).

92

Revenue is revenue. Just count it including all sources. Wealth in an accounting sense is assets less liabilities. State funding is not an asset. Endowment and tangible assets like land and property are assets.

Another thing, the endowments for universities are often given these days including quasi-endowments like funds that are held for medical center operations (i.e. not directly for any students). These can be a third or so of the total endowment or even more in some cases. You can see when they add these to their totals because they shoot up. Virginia Commonwealth University had an endowment of $400M a couple of years ago and started counting quasi endowments from the health system and now they cite $1.64B. This makes the endowment of schools like Princeton, which does not have a medical center, all the more remarkable.

93

Endowment is not a revenue stream. The income from endowment investments is a revenue stream. So is state funding. But so is income from other sources. You’re picking one income source, converting it into an endowment amount that you think would generate equivalent income, then adding the hypothetical value to the endowment. OK, that’s one perspective on institutional wealth. By accounting for financial resources in that way, yes, the ranking of some top state universities might improve. This is rather moot w.r.t. the USNWR ranking, because it doesn’t use endowment (or revenue) to rank colleges.

That seems to make sense of Michigan’s “S” number in its CDS S:F calculation. So it excludes medical and law students, but not graduate arts & science students. Some colleges seem to including the latter, while others don’t. The instructions should be clarified if necessary and all colleges should follow them consistently. If Duke added its arts & science grad students, its S:F would go up; it Michigan subtracted its a&s grad students, its S:F would go down. Either way, a fairly large difference apparently would remain in the ratios (which count for 1% of the USNWR ranking.)

94

IsaacTheFuture, universities use only 4%-5% of their endowment annually. So $300 million in state funding is the equivalent of $6 billion in endowment. So Michigan’s endowment, when compared to that of a private university, is not $10 billion, it is $16 billion, or $360,000/student. And UVa’s $150 million in state funding is the equivalent of $3 billion in endowment. So UVa’s endowment, when compared to that of a private university is not $7.5 billion, it is $10.5 billion, or $450,000/student.

Here’s another way of looking at it:

Revenues generated from endowment annually/student + state funding annually/student:

Brown University
Annual endowment spending: 5% of $3.1 billion = $150 million/9,400 students = $16,000/student

Columbia University
Annual endowment spending: 5% of $9.6 billion = $480 million/30,000 students = $16,000/student

Cornell University
Annual endowment spending: 5% of $6 billion = $300 million/22,000 students = $13,500/student
Annual state funding: $130 million/22,000 = $6,000/student
Total: $19,500/student

Duke University
Annual endowment spending: 5% of $8.5 billion = $425 million/16,000 students = $26,500/student

University of Michigan-Ann Arbor
Annual endowment spending: 5% of $10 billion = $500 million/44,000 students = $11,000/student
Annual state funding: $300 million/44,000 students = $7,000/student
Total: $18,000/student

University of Pennsylvania
Annual endowment spending: 5% of $10 billion = $500 million/22,000 students = $22,000/student

University of Virginia:
Annual endowment spending: 5% of $7.5 billion = $375 million/23,000 students = $16,000/student
Annual state funding: $150 million/23,000 students = $6,500/student
Total: $22,500/student

Like I said, financially, very little separates elite public universities from elite private universities.

“As for the academic staff, tk21769 already handles that.”

Actually, tk was wrong. My response in post #89 is correct. Graduate students must be included. Only programs that are purely open to graduate students, and their faculties, such as Law or Medical schools, should be omitted from the calculation. There is no other way of interpreting the method of calculating student to faculty ratio. Those that omit graduate students from their calculation are lying. Leaving graduate students out of the student to faculty ratio would suggest that graduate students do not take up any of the faculty’s time or resources. We all know that it not the case. In fact, graduate students are arguably more time intensive than undergraduate students.

Furthermore, looking at a university’s overall student to faculty ratio makes little sense. Usually, universities with large business and engineering programs will suffer in this department. It makes much better sense to break it down by college. I did so in post 84. Duke’s student to faculty ratio in its college of arts and sciences (Trinity) is 11:1, compared to the student to faculty ratio at Michigan’s college of arts and sciences (LSA), which is 14:1. And Duke’s college of engineering has a student to faculty ratio of 17:1 compared to Michigan’s 20:1. Do you honestly think there is a significant difference?

Listen, if the notion that public elites are in fact equal to private elites does not sit well for you, I will cease this discussion. I do not wish to upset you. But if you want to face the facts, then consider them as they are, not as you wish them to be.

“But no statistical evidence suggests that either are superior to Duke, UPenn, Columbia, Cornell, or any other similar elite private university.”

IsaacTheFuture, I never said that Cal and Michigan are in any way superior to schools like Cornell or Duke or Penn. I have clearly stated that they are all peers. Sadly, it is you who is suggesting that Cornell, Duke and Penn are superior to Cal and Michigan.

95

Isaac, having a medical school / system makes a huge difference in overall finances, particularly on research funding. A significant amount of research funding comes from sources like the NIH, and a lot of it goes to universities with medical schools and research hospitals. Look at the sources here: http://www.bestcolleges.com/features/colleges-with-highest-research-and-development-expenditures/ . On a per student capita basis, medical schools are typically among the best endowed schools at universities. Many of these can be endowments for research positions, though.

Berkeley has a pretty close relationship with nearby UCSF, which explains why they never added a medical school.

While having a medical center and medical school can add a lot of size and revenue to an institution, I don’t see that this directly benefits the undergraduate programs in most cases.

Regarding Duke vs Berkeley, it is a complex comparison. Berkeley is much stronger than Duke in many non-professional graduate areas. They are one of the top PHD schools. Duke is not at that level. Duke has a medical system and school, which skews the comparisons on finance. If you combined Berkeley and UCSF, it would likely look different. Duke probably gets significantly more net tuition and fees than Berkeley, which is not offset fully by state appropriations. In typical private school fashion, they use that money to attract bright students with merit scholarships. This mechanism is not available to public schools to the same extent.

96

“Lying” is a strong charge. If Duke and other private colleges are doing that, they aren’t going to great lengths to cover their tracks. Yale, Duke, Brown, Vanderbilt (the ones I’ve checked just now) all clearly show the number they’re using in the S:F numerator. But I agree there is a problem if schools aren’t all using the same method. I even agree it makes sense to include grad students in the “S”. So it would be interesting to add, say, the 2,230 Brown University grad students (minus the ones in “stand alone” programs) to the 6308 they reported for “S”. If every one of the 2,230 counts, the S:F would go up from 7:1 to 9:1. I’m not sure how we get to all the “stand alone” counts, though.

As for the spending numbers in post #93, Brown’s $16K, Michigan’s $18K, etc., wouldn’t nearly account for all expenditures. Some of these schools appear to be spending $50K per FTE, or more just on instructional spending (according to IPEDS numbers I’ve seen).

97

Alexandre,

I just don’t agree with your conclusions or methodology. You are essentially saying public schools get a subsidy from the state. We know that. But it misses a key point. When you take all sources of revenue, elite privates have an advantage in that they get significantly more in net tuition and fees, which is more than enough to overcome the state subsidy. Since they can charge more, they can do price discrimination. They charge good, rich students close to full tuition. They are willing and able to pay it. They use the surplus to offer better financial aid for poorer students (which is supplemented by Pell Grants, enabling the privates to charge a lot), and more importantly from a prestige perspective, to offer merit aid to get the best students. Elite publics, at least at the undergraduate level, can’t play this “game” as effectively, because they can’t do as much price discrimination because tuition is effectively capped. So the private school has more levers to optimize its enrollment stats, reflecting its elite status.

Note that elite public graduate schools are now operating as privates. UVA Law and Business essentially have private market tuition rates, as does Michigan. The economics I described above at the undergraduate level also explains why public schools are so keen to get more out-of-state students, who can be charged closer to market rates.

Also note that at some non-elite private schools, they are not able to play the price discrimination game. They are having to give bigger and bigger discounts just to fill spots. There is also an interesting Malcolm Gladwell podcast about how difficult it is for Vassar to maintain its standing because it uses comparatively more of its endowment and revenue to make the school more affordable. Its direct competitors are using more money to drive the amenities that wealthier students (who pay full freight) expect. This is putting Vassar in a tenuous position, despite its good intentions.

Playing around with endowment numbers is also misleading. You should never assume all of the money is distributed evenly, particularly to undergraduates. Most of it is restricted. For UVA, about 13% of the endowment is associated with the College of Liberal Arts and Sciences, which has 56% of all students. Even within that, a significant amount is restricted to specific uses. This applies to Ivy League schools as well. 80% of Harvard endowment is restricted to the owning school.

As an aside, using your methodology, UNC Chapel Hill may be more elite than UVA or Michigan, since they had been getting over $20K per student from the state. In return, though, the state expects them to keep tuition low.

98

In my copious free time I’ve been looking through the CDS section I2 numbers reported by some of the USNWR top 20 universities. They appear to be using at least three different methods to report S:F ratios:

  1. Some appear to be counting undergraduates plus a fraction of graduate/professional students. In other words they use approximately the same method Michigan does. JHU, Georgetown and Notre Dame seem to be doing it this way; they report 12:1, 11:1, and 10:1 ratios respectively.
  1. Others appear to be counting all full-time undergrads plus 1/3 of all part-time undergrads (but no grad students). At least 1 other T10 private school does it this way in its 2015-16 CDS, to arrive at an S:F ratio of 6:1.
  2. Still others appear to be counting only degree-seeking full-time undergrads plus 1/3 of degree-seeking part-time undergrads. At least 1 T10 private school does it this way in its 2015-16 CDS, to arrive at an S:F ratio of 6:1.

The T20 private schools that are using method 2 or method 3 may not all have enough arts & science grad students to drive their S:F ratios higher than the 10:1 to 12:1 range, if we included those students in the S:F. I suspect most (maybe all) of them would still be lower than Michigan’s 15:1. Significantly lower? Maybe not.

S:F only counts for 1% of the USNWR ranking. Still, I can understand why someone might conclude that a couple of variations like this poison the well for other CDS and USNWR data (even if some colleges are only being careless with instructions, not deliberately lying.)

99

I don’t understand the emphasis on revenue per student. At the high end, that’s mostly a measure of research dollars, which might be useful for measuring contributions to the public good or maybe, the quality of graduate education. I don’t see it as being useful for ranking undergraduate programs.

100

To ad on to what IzzoOne stated - in that you are using a flawed and biased methodology - many of the public schools do not have a surplus of revenue to spend on students, and even sometimes run a deficit. Take Ann Arbor, for example.

http://obp.umich.edu/wp-content/uploads/pubdata/budget/greybooksummary_fy17_allcamp.pdf

When you scroll down to page 21, you see that they are running a $84.9 million dollar budget surplus; however, NONE of that is coming from the general $1.94 billion general budget (in which they broke even with spending) or the $196 million designated fund (in which they broke even). They do have a $15 million surplus, but those are restricted funds. The other $69.9 million dollar surplus is coming from their health center; but even when you examine those numbers on page 24, you see that the medical school is in fact running a $113.6 million deficit, and are only saved from the $181.8 million dollar profit that the hospitals are making.
Even if we ignore the medical school deficit, and only look at the budget plus the $15 million surplus, we see that there is $74,703 for every student (Now this is ignoring that much of that money has to go salary, renovation , and other non-student related fees)

https://finance.duke.edu/resources/docs/financial_reports.pdf

A school like Duke on the other hand is running a $212 million surplus. Moreover, removing their patient service revenue that they get from their medical center (like we did for UM), we see that Duke is making a revenue of $185,074 for every student, which is more than twice of what UM is making, even though they receive state funding.

Even if we look at academic R&D Expenditure/Student (won’t include Berkeley due to lack of medical school, though I will note that in the areas where Berkeley shares with many of the privates, it is still lacking behind, even in engineering):

UMich: $30,620
Duke: $69,344
UVa: $16,976
Cornell: $40,326
Columbia: $31,070
UPenn: $34,735

UMich stacks up to UPenn, Columbia (still less), but lags behind Duke & Cornell; and it’s the same with other schools, such as Yale (which is comparable to Duke’s), JHU (which is in the 6 digits), MIT (6 digits, IIRC), Harvard, and even Princeton (despite having no medical school).

Even in engineering, Mich lags behind Duke, despite Duke not having a cost intensive course like Chemical Engineering.

In any financial statistical category you want to look at - endowment, budget, r&d - UMich, UVa, Cal, & many of the other top public lag behind their Private counterparts. No matter how you slice it, they simply can’t keep up financially. Now, does this mean they’re inferior schools? Not necessarily, or at least not purely because of financial reasons; for example, despite having less money to spend on its engineering school, outside of Biomedical Eng., UMich is a far better engineering school then Duke (though Duke’s is still great). However, FINANCIALLY speaking, Duke, and many of it’s top private peers, are simply wealthier than their public counterparts. It’s a benefit of being an elite private school. No need to be defensive about it.