Well, here is an answer:
http://www.improbable.com/airchives/paperair/volume7/v7i3/angels-7-3.htm
I’m sure we got too deep in the weeds, but the point was worth making in my view: Research oriented universities often make undergraduates pay a significant amount of their tuition and fees to subsidize research and graduate programs, which usually does not directly benefit the undergraduates.
In the case of the University of California, the analysis showed only 42% of what the university system said was devoted to undergraduate education actually was. The rest was used to subsidize research and graduate studies, and this amount per student was actually less than tuition and fees alone. So much for a true state subsidy of undergraduate educations. Undergraduates are actually doing the subsidizing, and a lot of them go into debt to do it.
IzzoOne, it is difficult to gauge exactly how much impact a university’s financial resources and spending have on undergraduate education, particularly in the case of research universities…obviously not in the case of LACs. But all major research universities operate in a similar way, so comparing a school like Michigan to Cornell or Northwestern etc…is reasonable since they all use their resources in a similar way.
As such, I appreciate that you took the time to look at another major research university and noticed that it too has similar spending patterns and practices. That is what I was saying all along; not that Michigan is exemplary, but that it is no different from other major research universities.
Here’s a breakdown of Cornell, Michigan and Northwestern, three schools I think share much in common. It should be noted that Cornell has 14,500 undergraduate students and 7,500 graduate students (22,000 students total), Michigan has 28,000 undergraduate and 15,500 graduate students (43,500 students total) and Northwestern has 8,500 undergraduate and 12,000 graduate students (20,500 students total).
REVENUES:
Net tuition (including graduate programs)
Cornell: $560 million
Michigan: $1.15 billion
Northwestern: $560 million
Government and non-government sponsored programs:
Cornell: $600 million
Michigan: $1.15 billion
Northwestern: $600 million
State Appropriations:
Cornell: $150 million
Michigan: $300 million
Northwestern: N/A
Donations:
Cornell: $170 million
Michigan: $130 million
Northwestern: $220 million
Income from endowment and other investments:
Cornell: $300 million
Michigan: $330 million (Michigan uses endowment income more conservatively, and sustainably, than most universities, who usually us 5% of their endowment annually, while Michigan caps at 3.5%)
Northwestern: $420 million
TOTAL REVENUES (not including medical centers):
Cornell: $2.5 billion
Michigan: $3.75 billion
Northwestern: $2.15 billion
As I have said before, when it comes to medical research, undergraduate students benefit very little, as it is almost exclusively confined to professors and medical students. But when it comes to research in fields such as Engineering, Biology, Chemistry and Physics, undergraduate students benefit directly as most undergraduate students in those fields actually work closely with faculty on research, and many even have their work published.
Proportionately, those three schools seem to have relatively similar and proportional revenue and expenditure streams. So while it is true that Michigan spends 2-3 times more in research from institutional funds, Michigan’s funds are twice as large as NU and Cornell to begin with. Also, one should not ignore economies of scale and cost of living allowances. Don’t you find it strange that when it comes to the USNWR financial resources rank, Cornell and Northwestern are both ranked among the top 15 but Michigan does not crack the top 40?
I appreciate your concern that only part of a research university’s resources are devoted to undergraduate students. I admit to it fully. My only request is that public universities be not accused of being less undergraduate friendly than private research universities. I think it is high time that private research universities admit that they too prioritize research and do not devote most of their resources to undergraduate students. Sadly, that is not the case. They continue to insist on publishing 3:1 or 5:1 or 7:1 student to faculty ratios, when in reality, their ratios are closer to 11:1 or 12:1, and they resort to far more creative accounting practices in order to report the dollar value they spend in undergraduate students.
You find Michigan’s $22,000 figure suspicious, and that is fair enough. Like I said before, I cannot prove how much Michigan spends on undergraduate students. But what of major research universities that claim that they spend $50,000 or more per student? Do you believe that in their case, the figure is even remotely accurate?
Well, perhaps we can agree on more this time. I was largely using Michigan as an example. My larger point was that research universities can divert a substantial amount of money from undergraduates to subsidize research and graduate studies. I’ll go back again to picking on someone else. For the University of California System, $18K was the published amount of spending per undergraduate for education per year. $7,500 was the actual amount, and it is actually less than tuition and fees. That is a huge difference. (Sorry to keep citing that one but I think it makes the point.)
One of the reasons I do not think $22K is accurate again is because the $500M spent on institutional funded research should be excluded. Institutions include this in their total, but is misleading because it is largely a transfer from undergraduate study to research. That is one of the points in the article I pointed to. Other schools are substantially overstated as well (e.g. UC System). Cornell, which you mentioned, spends $234M in institutionally funded research, or $10,600 per student. And it wouldn’t surprise me if this came disproportionately from the $14K undergraduates.
Unfortunately, almost all the stats on national research universities are skewed and misleading. I agree with you on student faculty ratios. They include faculty that don’t teach undergraduates. Furthermore, if one is looking at an institution, they should look at how many classes are taught by faculty rather than graduate students, and by faculty that are full time. Also, they should look at whether faculty are actually evaluated on how well they teach. Everyone always focuses on overall resources, but actual commitment to undergraduate education is likely more important. The sad trend is universities are spending more and more (and students are going more and more into debt), but faculty are teaching less and less and the teaching load is being outsourced to non-tenure track and graduate students.
I strongly disagree with your assessment that the institutionally funded research significantly benefits undergraduates. I cited a quote from a former Hopkins and University of Florida president that provides a similar viewpoint. Many of the highest performing schools in terms of outcomes (I gave the example of Washington and Lee) do not do it at all. Furthermore, if you are looking at PHDs production in science as a proxy (the NSF list), you will see that schools that do not do substantial transfers via institutional funding of research are very, very well represented. Institutions do not have to do this transfer to provide opportunities to undergraduates.
You probably don’t believe me, but I don’t have an axe to grind with the University of Michigan. I think the overall higher education system is too expensive and wasteful, doesn’t provide accurate information, and does too much cost shifting. And people looking seriously at colleges should evaluate the commitment of institutions to undergraduate education.
@IzzoOne , proof of the pudding is in the eating! My D2 is a Junior at GTech. She has at least 3 research opportunities for the upcoming Spring Semester. The one she is tending towards is one in the ECE Department, as though she is majoring in MechE, she wants to gain more experience in the area of Mechatronics! She says that similar opportunities are easily available for those that show interest at GTech.
IOW, reality beats all the stats thrown out by various sources for spending per undergrad at research univs!
Mechatronics. That’s a new one for me! Glad to hear she’s got great opportunities at Georgia Tech.
So @Alexandre, I didn’t answer your questions directly, so I’ll try to here. I agree that is is difficult to figure out how much is really devoted to undergraduates. Unfortunately, universities don’t really want people to know. What they do is combine undergraduate, graduate, and institutionally-funded research into the instruction bucket. You have to separate those out to have a true idea. Many favor graduate and institutionally-funded research. Because we are unlikely to get good numbers, I’d suggest looking at things like: 1) Do undergraduates have the chance to do faculty-guided research? 2) are teaching faculty evaluated based on teaching quality and effectiveness? 3) are classes taught by non-tenure and graduate students?, etc.
To your points, I think cost of living definitely matters, but you do get a class of professors who want to be in an area (take for instance UCLA) for factors they value beside cost of living. Also, there are schools located in areas where there are many outside opportunities to have additional outside income, particularly in certain fields. An example of this is NYU. Regarding student to faculty ratios, I completely agree that schools with lot of graduate and research-oriented faculty often provide misleading numbers that include those. I suspect the class size ones may also be manipulated, (e.g. not much difference between a class with 19 students and 21 students, but they may show up in different buckets.)
So, to a comparison of Michigan, Cornell, and Northwestern. If we use the IPEDS report and focus on Instruction, which I think is the most relevant area, we know that it actually includes (per a guideline that has been supported by NACUBO, the National Association of College and University Business Officers), it combines undergraduate, graduate, and institutionally-funded (or non-sponsored) research. They are pooled in this report. A cost accountant or a business consultant would look at this differently, and that is what the UC System report I mentioned did. We don’t really have any way to get at the graduate or undergraduate breakdown unless the universities provide us more data. But we can back out the non-sponsored research, which is really the university using its funds (which are accounted for as instruction) to pay people to do research. (There was an example of this in the NPR article on Duke.). If we do that, we get: Michigan – $22,728 Instruction (from IPEDS - $11,302 Institutional Research (from NSF) = $11,426 per student for real instruction (both graduates and undergraduates); Northwestern – $37,339 Instruction - $8,278 Institutional Research = $29,061 for real instruction; Cornell – $23,818 Instruction - $10,515 Institutional Research = $13,303 per student real instruction. So, to your point, Michigan is quite close to Cornell (an Ivy school) in this calculation, but trails Northwestern by a fair amount. To throw Duke in, it would be $50,756 Instruction - $9,573 Institutional Research = $41,183, which is quite a bit more.
Note again that this does not separate out graduate and undergraduate. The UC System report did that by looking at time spent by function, but we don’t have that data. (The California assembly has added language saying they want the costs to be reported separately going forward. I think more should do that.) In general, I think schools spend quite a bit more on a per student basis on graduate students than undergraduates. Professional schools (like Law, Medicine, and Business) have higher paid professors that often don’t do undergraduate teaching, and Masters and PhD students typically also have more time with faculty.
It likely that the data I put in for Cornell is distorted by having Cornell separated into Cornell University and Weill Cornell in the IPEDS data. I didn’t pick that up originally. Medical schools drive the numbers up. The average instruction cost there is listed at $193K per FTE. Web site says they have more than 400 students.
“To your points, I think cost of living definitely matters, but you do get a class of professors who want to be in an area (take for instance UCLA) for factors they value beside cost of living.”
Obviously, different profs will seek out different settings. I do not think Ann Arbor suffers in this department though. It is one of the more sought out college towns in the country, and its ability to attract and retain leaders in every field has never been a challenge. California schools do have a tendency to attract retiring Nobel Prize winners though. 
As I said earlier, I find the discrepancy between Cornell/Michigan and their peers (Columbia, NU, Penn etc…) highly suspicious, and alarming. Clearly, they are not calculating spending the same way. As I already demonstrated, they all have approximately equal revenues on a per student basis. It is mathematically impossible for them to spend 2-3 times more per student.
“In general, I think schools spend quite a bit more on a per student basis on graduate students than undergraduates. Professional schools (like Law, Medicine, and Business) have higher paid professors that often don’t do undergraduate teaching, and Masters and PhD students typically also have more time with faculty.”
I agree, which is why I think schools like Cornell and Michigan, where only 35% of the students are graduate students, actually spend more, relatively speaking of course, on undergrads than schools where 70%-80% of the students are graduate students.
Great discussion guys. Lots of great data provided to the forum.
Personally, I feel the detail is just not provided by the schools to adequately sustain a position. But is does make for a nice discussion.
All this leading us to another college ranking----Professors:
http://www.bestcollegereviews.org/50-schools-best-professors/
Interesting side note, it was very interesting and (cool for my son) to see the professors at Dartmouth walking or biking to class since they all seem to live right there on campus since it’s such a small town/community.
@moscott, It is too bad the best professor ranking only has private schools other than the service academies. I’d like to see some public schools in there.
@Alexandre, I suspect the issue with Cornell has to do with the IPEDS data having Cornell University and Weill Cornell Graduate School of Medical Sciences separated. If you combine them, the average for IPEDS Instruction goes from $24K per student to $31.5K per student (Weill Medical is $193K per student for Instruction). This is quite a bit closer to Northwestern. If you back out non-sponsored research, it would be $21K per student per year. I’d have to look more into Michigan, but I do see that Northwestern is about 22% above Michigan in overall “resources” based on the number you provided, and about 57% according to the IPEDS data.
@Alexandre, here is what I see for 2015 from the sources of funds instruction MIGHT come from:
Net Tuition and Fees:
Michigan Ann Arbor – $989M (Estimated – see note below)
Northwestern – $568M
State:
Michigan Ann Arbor – $299M
Northwestern – $0
Endowment and other Investments:
Michigan Ann Arbor – $332M
Northwestern – $472M
Gifts:
Michigan Ann Arbor – $133
Northwestern – $216M
Total:
Michigan Ann Arbor – $1,754M
Northwestern – $1,255M
Students:
Michigan Ann Arbor – 43,651
Northwestern – 20,995
Per Student:
Michigan Ann Arbor – $40,171
Northwestern – $59,900
Northwestern is 49% higher than Michigan based on this analysis. According to IPEDS it is 64% higher for its instruction category.
Regarding net tuition at Ann Arbor, what I found is total system net was $1,146M for 2015, and gross was $1,507M. I used proportion of gross revenue to estimate that net revenue at Ann Arbor was $989M.
Again, keep in mind instruction bundles undergraduate, graduate, and institutionally funded research into one category.
Seems accurate IzzoOne. However, the Ann Arbor campus generated $1.1 billion in net tuition in 2015. $1.3 billion in tuition minus $200 million in financial aid. It won’t make a big difference mind you, but I value accuracy. It should also be noted that a far larger portion of tuition revenues at Northwestern is generated by graduate students. While Northwestern has roughly one third the number of undergraduate students that Michigan has (8,500 vs 28,000), it has almost as many graduate students (13,000 vs 16,000).
Regardless, my point is that there is no way Northwestern can afford paying 250% more on a per student basis than Michigan. It does not have the resources to do so without driving itself into the ground. Those two universities are coming to their respective numbers using completely different methods. I can see how Michigan benefits from economies of scale and can cut costs by a significant margin, but not by 250%. And Northwestern is mild compared to other schools that claim spending $50k or more per student. Again, one should only look at revenues when comparing research institutions. Expenditures are not telling in the case of research universities because there are far too many streams to account for. But it is safe to say that virtually all top universities spend a similar amount on undergraduate students once you account for cost of living and economies of scale. It would not be economically viable for Northwestern to spend 250% more than Michigan on its students.
The link below breaks it down according to campus. If you look under the Ann Arbor campus, you will see the figures I refer to in pages 1 and 37.
http://obp.umich.edu/wp-content/uploads/pubdata/budget/greybookdetail_fy16_allcamp.pdf
The IPEDS indicates Northwestern spends 64% more per student for Instruction, not 250%. That seems within reason given the analysis above. Again, and this was my main point, that IPEDS total, if accurate, will still bundle undergraduate, graduate, and institutionally (internally) funded research, and is likely not to be an accurate indicator for undergraduate students for research universities. We can back out institutionally funded research which is 49% and 22% of the instruction totals at Michigan and Northwestern respectively. We can’t separate graduate and undergraduate unless the institution gives us more information. You can see the impact of graduate programs in Cornell, though. IPEDS separates Cornell University and Weill Cornell, the medical college. Weill Cornell spends, according to IPEDS, $193K per student in instruction while Cornell University (without Weill) spends $24K. That is 8X as much! If you combine them, it brings Cornell + Weill up to $31.5K per student, a 31% increase.
@Alexandre, I had looked at the report you cited, but never saw a net tuition amount. There is a net tuition for the system in another report I saw, and that is what I used for the estimate.
“The IPEDS indicates Northwestern spends 64% more per student for Instruction, not 250%.”
That seems fair. Michigan does benefit from economies of scale after all, and like I said, faculty at Michigan are not paid as much because cost of living is slightly lower. The 64% difference considering those to factors is pretty much negated.
“I had looked at the report you cited, but never saw a net tuition amount. There is a net tuition for the system in another report I saw, and that is what I used for the estimate.”
Can you share the link to that report. I trust you, but perhaps it is not complete. It seems unlikely that Michigan-Ann Arbor raises $1.3 billion in tuition, but that the net tuition revenue is only $980 million.
This was it: http://www.finance.umich.edu/reports/2015/ Page 9. Only reference to “net” I could find.
Main sufferers are the top non-ivy private schools that aren’t as well known by everyone and therefore suffer on certain lists. Same results help the top publics with this methodology.
Anyways, I think most can agree that most people (costs or geographical location compared to themselves) would take Carnegie Mellon over half the schools listed above them - not that selectivity or yield is the perfect test of how “great” a university is for its students either :).
@Jpgranier You seriously think UCLA’s acceptance rate is inflated? They receive more applications than anyone due to the vast amount of in-state applications that they receive (many times from students who would otherwise not apply but do so because they apply to some other UC school and its easy) as well as the amount of international and out of state applications that they receive. The tuition at UC OOS is still less expensive than the tuition of many private schools.
Also please explain what you exactly mean by “inflated” acceptance rates? Do you mean that the school is actually more selective and difficult for admission than the acceptance rates imply? For a school’s acceptance rates to be inflated in these respects would mean that it has a very self-selective admissions process where many unqualified students do not apply - a process that occurs much more at private schools than public ones. The average SAT score would imply otherwise about UCLA having an “inflated” acceptance rate.
(acceptance rates are pointless anyways)
- When I said inflated I meant that their acceptance rates are higher in number than the should be. I was trying to say like 40% vs 20%. I can see how it came out wrong. When I posted this, the state that came to mind was the Texas school system. They have an automatic acceptance system, so there are plenty of unqualified (test score) applicants that are able to get in due to being the top 7%.
- I didn't call out UCLA, so I don't see why you're obsessing over it. I was saying most public schools, imo, should have lower acceptance rates because they have a smaller applicant pool due to instate/out of state pricing.
- Don't understand why you're tagging me. I posted that two months ago. I had to search the thread because I didn't even remember posting that...