Net university spending per student

College Search & Selection
1

<p>I was playing with IPEDs and Excel today, and made some calculations to determine the amount of money that a variety of colleges and universities spend on their students. I calculated the total amount of money in student spending by combining 3 IPEDs categories: Instructional spending per student, Academic Support spending per student, and Student Services spending per student (note that this does not include Research spending, among other areas, as the vast majority of such spending, though not all, does not directly benefit undergrads. A few other such categories were also left out, including Public Service spending, and Institutional Support spending). All of these numbers include graduate and professional school students in both the money spent numbers and the total student numbers, so they may not perfectly reflect undergrad spending.
I then decided that another important component was how much the student must spend to receive these services. I therefore subtracted the tuition and fees revenue per student from the previously calculated number. Note that the tuition and fees number is not simply the list cost. Financial aid provided by the school does not result in revenue, so the tuition and fees numbers are all less than the amount actually charged by the university. These numbers would be different for someone paying the list price. Moreover, for public universities, these numbers reflect the ratio of in-state to out-of-state students. Since out-of-state students pay more tuition, having a large number of such students will inflate revenue, hurting a school's numbers in this listing. And, obviously, the real ranking for an individual would depend on whether a given public school is in-state or out-of-state.
Without further ado (in rank order, institutional spending per student on students minus the amount of tuition and fees received per student, in dollars):
Yale University 95549
California Institute of Technology 87892
Washington University in St Louis 59634
Stanford University 55399
Johns Hopkins University 54346
Harvard University 50878
University of Chicago 42313
Princeton University 38327
Dartmouth College 37888
Columbia University in the City of New York 34549
University of Pennsylvania 33723
Rice University 26496
Williams College 24618
University of California-Los Angeles 24218
Swarthmore College 21068
University of North Carolina at Chapel Hill 20503
Pomona College 20366
Duke University 19857
Northwestern University 18307
Cornell University 15468
University of California-San Diego 15458
Amherst College 13783
Harvey Mudd College 13155
Brown University 12908
Middlebury College 11740
University of California-Berkeley 11521
University of Wisconsin-Madison 10148
University of Michigan-Ann Arbor 7199
Georgetown University 6461
University of Virginia-Main Campus 5154</p>

<p>Interesting results. One thing that particularly surprised me was the amount of variation among schools in the UC system. UCLA is spending far more per student than Berkeley or SD. Perhaps UCLA is really the UC school where one should go to get his or her money's worth.
Another thing to note is how high WashU is. Despite the amount of bashing it gets on these boards, it clearly invests a lot of resources in its students (and this does not include research money, so its medical research is not affecting the results). Some of the CC favorites don't do so well. Brown, an Ivy, does worse than all but 6 other schools out of the 30 I looked at.
One note about Michigan. It had a particularly high tuition revenue per student number for a state university. I believe it has many more out of state students compared to the other public universities I looked at, so its ranking would rise more for an in-state student than any of the other publics.
Obviously I can't say for sure that universities aren't reporting number differently from each other. I will be the first to admit that the magnitude of some of the differences is somewhat surprising (particularly Yale/Caltech vs the rest of the field). On the other hand, these numbers are from a federal database, and one would hope that institutions are accurately reporting data to the government.
I'm willing to calculate the numbers for more schools if people are interested (I decided to focus first on the ones that get the most interest on these boards). But don't expect me to respond immediately. I'm also open to suggestions to improve the methodology.</p>

2

<p>Data for more schools (should now include entire USNWR top 30 National Universities and top 15 LACs). I will also state where each new school would place in the original list.</p>

<p>Wake Forest University 68863 (3rd, behind Caltech, in front of WUSTL)
Vanderbilt University 45485 (8th, behind Harvard, in front of Chicago)
Emory University 24998 (15th, behind Rice, in front of Williams)
Wellesley College 21155 (18th, behind UCLA, in front of Swarthmore)
Grinnell College 20425 (21st, behind UNC, in front of Pomona)
Carnegie Mellon University 16919 (25th, behind Northwestern, in front of Cornell)
Bowdoin College 14260 (28th, behind UCSD, in front of Vassar)
Vassar College 14251(29th, behind Bowdoin, in front of Washington and Lee)
Washington and Lee University 14027 (30th, behind Vassar, in front of Amherst)
Haverford College 13673 (32nd, behind Amherst, in front of Harvey Mudd)
Claremont McKenna College 12853 (35th, behind Brown, in front of Davidson)
Davidson College 12373 (36th, behind CMC, in front of Middlebury)
Carleton College 11685 (38th, behind Middlebury, in front of Wesleyan)
Wesleyan University 11569 (39th, behind Carleton, in front of Berkeley)
University of Southern California 9673 (42nd, behind Wisconsin, in front of Tufts)
Tufts University 9486 (43rd, behind USC, in front of Notre Dame)
University of Notre Dame 8060 (44th, behind Notre Dame, in front of Michigan)</p>

<p>just realized I missed MIT
Massachusetts Institute of Technology 59061 (5th, behind WUSTL, in front of Stanford)</p>

3

<p>
[quote]
Perhaps UCLA is really the UC school where one should go to get his or her money's worth.

[/quote]
</p>

<p>Your "perhaps" is important!</p>

<p>Not all expenditures enhance education equally--one school might spend more just to offer what its peers can provide with a lot less spending. For example, a college which is located in an urban area may have to expend much more on campus security than a peer which is located in a rural or suburban area. A campus which is spread out may need to spend more on transportation services. A school which is less prestigious may spend much more per student on admissions. All of these things will make the per-student figure higher, but wouldn't necessarily mean the education is better.</p>

<p>Program mix also has a big impact. If a school has programs that tend to be expensive to provide (engineering, nursing, etc), their expenditures per student will skew higher.</p>

<p>And finally, if athletics isn't run as an auxiliary enterprise, that can be included in student services, too.</p>

<p>Obviously, these things won't account for all the differences we see between schools, but it muddies the interpretation a bit and it makes it hard to say you're definitely getting "more for your money" on any campus where per student spending is higher.</p>

4

<p>Hoedown, you are, of course, correct. I consider these numbers interesting more than definitive. Only the most rigid of economists would claim that the amount of money an institution invests in you is the right way to choose a school. But they do mean something, and are, IMO, more useful than some of the data that goes into the US News (though a form of these numbers appears in "financial resources").</p>

5

<p>Well, I think that's a top-10 list we can all live with. (jk)</p>

<p>I was wondering how spending on athletics would fit into this --- is it counted in your metric or not? And did you "catch" this bug from collegehelp?</p>

6

<p>Intramural athletics spending counts. Intercollegiate athletic expenses seem to count to the extent to which they are funded by the university, rather than self-sustaining. Thus, the large, revenue making football and basketball programs at places like Notre Dame, USC, Stanford, Berkeley, Michigan, UCLA, etc, do not count toward spending. It would be nice to be able to separate out intercollegiate athletic spending, but it is part of a category that is very important (student services, which includes all university funding of extracurricular activities, student organizations, and student health services). Eliminating the category entirely has only a small effect on the final numbers at the top (Yale drops to 2nd, Columbia and Penn go up a few spots, Williams drops a long ways, UCLA rises a few) but a rather large effect on the bottom, where student services makes up a much larger proportion of the total spending (some highlights include Georgetown falling to last, a big drop for Bowdoin, Washington and Lee has a big gain. Wesleyan, Berkeley, and Wisconsin all benefit).</p>

<p>I may have caught the collegehelp bug, but I think I've now covered the two areas that intrigue me the most, so this will probably be my last thread of this sort.</p>

<p>
[quote]
Well, I think that's a top-10 list we can all live with. (jk)

[/quote]

Except for Wake Forest, the top-10 schools in this are all in the top 20 in US News, so there's some correlation (not that US News is a very good ranking). Though LACs seem to do poorly, as Williams, the highest, is 17th.</p>

7

<p>It's a good effort on your part. It should make for some interesting discussion come morning.</p>

8

<p>"All of these numbers include graduate and professional school students in both the money spent numbers and the total student numbers, so they may not perfectly reflect undergrad spending." </p>

<p>If that is the case, this ranking is pretty inaccurate.</p>

9

<p>UCLA is actually probably ranked too high, so that would make sense. The per capita spending consideration really hurts public schools.</p>

10

<p>The low ratings of the LAC's implies that this comparison, like many others, is interesting only among similar institutions. It also reinforces the idea that the spending is heavily determined by the size and nature of programs. The LAC's do not have huge, expensive, scientific research institutions. </p>

<p>Factoring in financial aid is an interesting characterization of institutional practice, but perhaps obfuscates as much as clarifies. Variation in the portion of students who get outside aid will be missed, as I understand the metric. It also diverts attention from the investment in individual students. Thus any given student has whatever level of support they actually get (0-100%), and the mean level of support for the other students has only a very indirect effect on their experience.</p>

<p>Among potential distortions may be the definition of FTE, which is full time students plus 1/3 of part time students. This inherently assumes that part-time students consume much less in resources than do full time students. To the extent this is incorrect, and varies across the colleges, the results will be misleading.</p>

<p>I suspect that these figures reflect a wide range in institutional organization that yields wide ranges of accounting reports. Thus some costs that are counted in student expenditures at college A would not be included at college B. For example, I doubt that, before considering aid, Yale really spends twice as much per student as do Harvard or Princeton. If it did, where is this money going? </p>

<p>The results also suggest that the figures are heavily influenced by the presence of graduate students. One striking example is the comparison of Harvey Mudd to Caltech. Mudd is as close as one can get to Caltech without the graduate programs, with similar size and make up of undergraduate populations. However, according to this metric, Caltech spends almost 7 times as much per student as does Mudd. What genuine difference in student experience could possibly explain this large a difference in expenditures?</p>

11

<p>svalbardlutefisk,
Many thanks for doing the work in creating this analysis. Obviously, you put a lot of time and effort into it and tried to create a fair result. </p>

<p>Re the actual analysis and what it means, I feel strongly that these results should have great value as a student looks at a college and tries to understand what his/her undergraduate experience is going to look like. Schools with strong academic reputations that also spend a lot of money to support undergraduates is the ideal undergraduate academic setting for a student. Students benefit more directly at these schools and this should not be underestimated. </p>

<p>Special recognition should be given to a few schools that do particularly well in your analysis yet have a little lower profile than the usual suspects. In particular, Wake Forest and Wash U rank 3rd and 4th and Johns Hopkins is 7th and Vanderbilt comes in at 9th. Their administrations should be applauded for making such a strong effort to support undergraduate education. </p>

<p>For the national universities, I took the liberty of separating out the private from the public universities. Here is the updated list using the numbers you crunched:</p>

<pre><code>PRIVATE UNIVERSITIES

</code></pre>

<p>1 Yale $ 95,549
2 Caltech $ 87,892
3 Wake Forest $ 68,863
4 Wash U $ 59,634
5 MIT $ 59,061
6 Stanford $ 55,399
7 Johns Hopkins $ 54,346
8 Harvard $ 50,878
9 Vanderbilt $ 45,485
10 U Chicago $ 42,313
11 Princeton $ 38,327
12 Dartmouth $ 37,888
13 Columbia $ 34,549
14 U Penn $ 33,723
15 Rice $ 26,496
16 Emory $ 24,998
17 Duke $ 19,857
18 Northwestern $ 18,307
19 Carnegie Mellon $ 16,919
20 Cornell $ 15,468
21 Brown $ 12,908
22 USC $ 9,673
23 Tufts $ 9,486
24 Notre Dame $ 8,060
25 Georgetown $ 6,461 </p>

<pre><code>PUBLIC UNIVERSITIES
</code></pre>

<p>1 UCLA $ 24,218
2 U North Carolina $ 20,503
3 UCSD $ 15,458
4 UC Berkeley $ 11,521
5 U Wisconsin $ 10,148
6 U Michigan $ 7,199
7 U Virginia $ 5,154 </p>

<p>One caveat that hoedown points out that I want to support is how the nature of the programs will have consequential impact on spending patterns, ie, academic programs like engineering and health related programs, even outside of their research activities, will carry a larger cost than those for something like history or languages. </p>

<p>There were a few surprises in here such as the fact that the first LAC (Williams at $24,618) was exceeded by 16 National Universities. Given the focus of the LACs on undergraduate education, I expected them to be a bit more competitive. Here is the data for the LACs that you listed:</p>

<pre><code>LACs

</code></pre>

<p>1 Williams $ 24,618
2 Wellesley $ 21,115
3 Swarthmore $ 21,068
4 Grinnell $ 20,425
5 Pomona $ 20,366
6 Bowdoin $ 14,260
7 Vassar $ 14,251
8 W&L $ 14,027
9 Amherst $ 13,783
10 Haverford $ 13,673
11 Harvey Mudd $ 13,155
12 Claremont McKenna $ 12,853
13 Davidson $ 12,373
14 Middlebury $ 11,740
15 Carleton $ 11,685
16 Wesleyan $ 11,569</p>

12

<p>
[quote]
The results also suggest that the figures are heavily influenced by the presence of graduate students. One striking example is the comparison of Harvey Mudd to Caltech. Mudd is as close as one can get to Caltech without the graduate programs, with similar size and make up of undergraduate populations. However, according to this metric, Caltech spends almost 7 times as much per student as does Mudd. What genuine difference in student experience could possibly explain this large a difference in expenditures?

[/quote]

The numbers are, of course, divided by total number of students, including grad students. So hopefully that does combat the problem somewhat. Though, admittedly, it is surprising that LACs do so poorly. But in order for say, that Caltech/Mudd difference to be due to grad students, the spending per capita on grad students would have to be 5 or 6 times as high as the per capita spending on undergrads. Still, separating by type of institution is a good way to use the data
On reason research universities may do well is higher faculty salaries, as faculty at major research powers get paid more than their LAC counterparts, but this alone wouldn't explain the entire difference.</p>

<p>
[quote]
The low ratings of the LAC's implies that this comparison, like many others, is interesting only among similar institutions. It also reinforces the idea that the spending is heavily determined by the size and nature of programs. The LAC's do not have huge, expensive, scientific research institutions.

[/quote]

Direct research spending was not included in the ranking. The LACs would have done far worse if it had been.

[quote]
"All of these numbers include graduate and professional school students in both the money spent numbers and the total student numbers, so they may not perfectly reflect undergrad spending."</p>

<p>If that is the case, this ranking is pretty inaccurate.

[/quote]

Not necessarily. Unless per-student spending levels are much higher for grad students than undergrads, it shouldn't have much of an effect. In fact, I hypothesize that large numbers of grad and professional students might explain why Harvard does so much worse than Yale in this metric - that size of grad programs may be working against it.</p>

<p>
[quote]
</p>

<p>Factoring in financial aid is an interesting characterization of institutional practice, but perhaps obfuscates as much as clarifies. Variation in the portion of students who get outside aid will be missed, as I understand the metric. It also diverts attention from the investment in individual students. Thus any given student has whatever level of support they actually get (0-100%), and the mean level of support for the other students has only a very indirect effect on their experience.

[/quote]

I agree entirely. I used the numbers I did because that was all IPEDs had. I could have gone through every institution to find its full costs, but that would have been an extremely time consuming process. If you want to either compile this data, or point me to a link where it is all in one place (so it doesn't take hours to put together), I would be happy to redo the numbers with the full price numbers.

[quote]
Among potential distortions may be the definition of FTE, which is full time students plus 1/3 of part time students. This inherently assumes that part-time students consume much less in resources than do full time students. To the extent this is incorrect, and varies across the colleges, the results will be misleading.

[/quote]

I don't think many of these universities have huge numbers of part-time students (particularly the privates on this list), so it hopefully is not having a big effect. And part-timers do consume somewhat fewer resources.</p>

<p>Thanks, hawkette, for separating the numbers. It may make them more useful. One note about Wake Forest: the reason it does so well is its spending on Academic Support which is at least 4-5 times higher than that of all but two other institutions (Harvard and Penn, which are 2-3 times as low). Perhaps they just report numbers differently, but it seems that something odd may be happening here, especially as that category is probably the least important to undergrads of the three I included (though it does have relevance).</p>

13

<p>I don't exactly understand what these numbers mean. Being a student who goes to the school which is listed dead last on this list in "money spent" on me, I don't see what I'm missing from a student who is at Yale.</p>

<p>A few things to keep in mind. I don't really buy this "how much they spend" per student idea, because it obviously depends on what the student is doing. I am a history and music major...and I'm sure the university has to invest less in me than someone who is a biomedical engineer, a surgeon in training, or some quantum physics grad student who gets to play with some $1 million computer.</p>

14

<p>Amazing post. Just wonderful, Sval. Now we have something that is extremely thought provoking. Students and parents looking for the best place for their children should consider your findings, or some variation thereof. Congratulations on doing a wonderful service for many, many people.</p>

15

<p>I think the problem with methodology is that many universities don't distinguish between certain types of research and instruction. For example, if a professor is compensated 120,000, the number isn't divided 60k research, 60k instruction. Chances are, all 120k is put into the instruction category. The research given by the OP might be only organized/sponsored research, not departmental research.</p>

<p>Lets look at Yale:
According to <a href="http://72.14.253.104/search?q=cache:mvuuYhyGUJwJ:www.yaleinsider.org/works-budg.html+yale+operating+budget&hl=en&ct=clnk&cd=1&gl=us&client=firefox-a%5B/url%5D"&gt;http://72.14.253.104/search?q=cache:mvuuYhyGUJwJ:www.yaleinsider.org/works-budg.html+yale+operating+budget&hl=en&ct=clnk&cd=1&gl=us&client=firefox-a&lt;/a>
(the original site doesn't work, so I looked at a google cache)</p>

<p>Yale's operating budget is 1.68 billion
Organized (sponsored) research accounts for 20%
Patient Care accounts for 15%</p>

<p>Subtract these sections, then the budget becomes 1.09 billion, divided by the total students (11,390), and the result is 95873 per student. If we account for the rise in budget over 4 years then subtract net tuition and fees, this number should be similar to what the OP got.</p>

<p>However, the biggest remaining section of the fees is "Instruction and Department Research" at 30%. The 95873 is only when I assume all of that is instruction, and none is for research. I think some or all of the schools listed put departmental research in the same area as instruction, which accounts for why research universities appear to spend so much more than liberal-arts colleges.</p>

16

<p>There will always be some concern with methodology in any analysis set. For example, I would like to see research dollars spent per undergrad student included, because I know that Hopkins places an enormous emphasis on it, relative to other institutions. Research on the undergrad level, as I have pointed out before, is a main focus of its educational philosophy.
But, understanding the limitations, in terms of the assessment that Sval presented, they idea is to look at the trend...the broad view. We can all get the idea, and draw general conclusions from it. No matter what the flaws...what's missing, what might have been included...still a very valuable piece of the decision-making pie.</p>

17

<p>I wonder if different schools do their accounting differently and therefore report their IPEDS data differently. I wonder if the same things are going into each of the three IPEDS categories that you used for each school.</p>

<p>Each of the three categories includes subcategories:
1. salaries and wages
2. benefits
3. operation and maintenance of plant
4. depreciation
5. interest
6. "all other"
Some of the above categories don't directly impact student's educational experience. I think professor salaries is most relevant.</p>

<p>When I took just instructional "salaries and wages" (excluding academic support and student services) and divided by enrollment, this is what I came up with for instructional salaries per student for the Ivies:</p>

<p>school, instructional salaries per student, enrollment, total instructional salaries</p>

<p>Yale University $36,945.31 11483 $424,243,005
Columbia University in the City of New York $25,625.21 21983 $563,319,000
Dartmouth College $15,356.92 5780 $88,762,993
University of Pennsylvania $14,785.39 23704 $35,0473,000
Harvard University $14,742.42 25017 $368,811,000
Princeton University $14,701.46 6773 $9,9573,000
Cornell University $12,120.02 19642 $23,8061,449
Brown University $11,945.65 8261 $98,683,000 </p>

<p>Yale and Columbia are much higher than the other Ivies. The other Ivies are roughly the same. Yale and Columbia might have a different way of defining who is considered "instructional". It creates a different impression than post #1.</p>

18

<p>collegeresults.org is a good source for a lot of college data in tabular form</p>

<p>For expenses, it reports the following data from IPEDS</p>

<p>
[quote]
Instructional Expenditures / FTE: “Instruction expenses” is a discrete reporting category. It includes expenditures for the colleges, schools, departments, and other instructional divisions of the institution and expenses for departmental research and public service that are not separately budgeted. It also includes general academic instruction, occupational and vocational instruction, community education, preparatory and adult basic education, and regular, special, and extension sessions. It also includes expenses for both credit and non-credit activities. It excludes expenses for academic administration where the primary function is administration (e.g., academic deans). Information technology expenses related to instructional activities if the institution separately budgets and expenses information technology resources are included (otherwise these expenses are included in “academic support”). (IPEDS)</p>

<p>Educational and General Expenditures / FTE : This is a broader category, which includes the instructional expenditures listed above, plus expenditures for research, public service, academic support, student services, institutional support, plant operation & maintenance, and scholarships. (IPEDS)</p>

<p>Student and Related Expenditures / FTE : This is an intermediate financial measure, including instructional, student services, and academic support expenditures. The specific formula was developed by the National Center for Higher Education Management Systems (NCHEMS). Student-related expenditures are calculated as (Instruction + Student Services + Academic Support*(Instruction/(Instruction + Public Service + Research))). (IPEDS)[/quote}</p>

19

<p>What is most stunning about these numbers is how few dollars some top schools seem to spend on their students. I am most surprised by Tufts, Cornell, Brown, Notre Dame and Georgetown. If these figures are anywhere in the ballpark, caveat emptor.</p>

20

<p>That's why I am cautious about the figures. The numbers for Williams, Amherst and Swarthmore show what it costs to give students about as good an undergraduate education as there is. Spending more, and in some cases, several times more, makes me suspect that the figures do not really reflect money spent on undergrad education.</p>

<p>If Yale is really spending anywhere near that much on undergrad education, where is the money going? It pays its professors well, but no better than Harvard. The facilities, financial aid, etc are quite comparable. So what does all this extra money buy?</p>

<p>One other concern- short term educational programs sponsored by the professional schools. Medical and business schools do a great business offering courses that range anywhere from 1 day to a couple of weeks on narrow topics. These are for working professionals and executives. Law schools may do the same thing, I don't know. </p>

<p>These programs generate quite a bit of expenses, as professors must be paid, halls rented, utilities, insurance, depreciation, etc are all charged to the effort. These programs never enroll undergrads, so they contribute nothing to undergrad education. Some are not even based at the main campus. The faculty, being at the professional schools, have little interaction with undergrads. So none of the expense enhances the undergrad experience.</p>

<p>However, as best I can tell, all of this expenditure would count as instructional expense. Since this requires professional schools, the LAC's are not in the game at all. The more LAC-like universities with few or no professional schools also would be out of the game. </p>

<p>Universities with the full range of professional schools would then compete for bringing students into these programs. Places in major population centers (Columbia, Harvard) and with big names (Columbia, Harvard, Yale), would show high instructional expenses. Divide this by the number of students, and institutions with large professional programs, and, by research university standards, relatively small student bodies, would show high expenses per student.</p>