Students, Money, and Scholarship.

Before reading further, I want to add some comments and disclaimers: 

1.   I used the Web of Knowledge to locate papers from a set of Universities. It is not an exhaustive study. 
2.  The publication data here was collected over the period November 2011 to March 2012 of so may be different now. 
3.  My ability to find every article is imperfect. In some cases, I could be off by a factor of 2. As an example, I found fewer publications in some cases than did the Leiden index. 
4.  The publication data I have is for a 20 year period ending in 2012. 
5.  I genuinely believe the data to be indicative and since it is presented on a log-log plot factors of 2 on the y-axis make very little change. If I have made any errors that fundamentally change the conclusions, please bring them to my attention. 
6.  The presentation of US traditional universities is sparse. The reason for this is that I find the annual financial reports nearly impossible to read. The UK equivalents get right to the point and are nicely summarised.

So here goes...

Many academics are suspicious of metrics. Usually, discussions of university metrics try to discern the best university in the world, the second best, and so forth. I really do not care about this. In honesty any university in the THE top 200 is a very good university. Further, many universities that you will not find ranked will allow you to get a very good education. Having been around the block a bit, I can honestly say teaching to students with the great test scores is easier than it is to those with lower test scores. Yet universities get hit in league tables for taking the latter. Low performance in school between ages 16-18 can mean many things and all too often it can mean location, family background, and insecurity rather than anything to do with ability as an adult.

What am I up to here? I am interested in the spectrum of HE, what it does, and how it can improve. As an analogy, there isn’t very much variation in elite marathon runners in terms of measurable physiology. There is by comparison much more in the general population, many of whom get by just fine.

One thing a university does is scholarship. Scholarship costs time and time is money. Scholarship is important because Universities are a “content” industry and generating the content that future generations will study is a necessity. In the absence of advancement and scholarship we become schools. The other thing a university must do is maintain its financial stability.

The data shown is a plot of two metrics: 

1.  The total money through the books in a year (2010/11) divided by the number of students. This is NOT the amount each student pays. Universities like Cambridge and Stanford have very large businesses augmenting income from research grants and student fees. 
2.  The number of papers I found when searching on the Web of Knowledge over a period of 20 years divided by the number of students in 2010/11.

Why these? The idea was to get a sense of student normalised income and scholarship. A university turning over £15K/student should be roughly comparable to any other university turning over the same amount. One might predict some economy of scale for larger institutions. As more income is obtained more scholarship would be expected.

Let’s look at the data.

Is there anything here of interest?

1. The UK seems to follow a single trend all the way from the Open University (lowest income per student) through to Cambridge and Imperial. As income per student rises, so does scholarship, and it does so commensurately across all that I looked at. This is quite remarkable and goes against the rhetoric of UK university groupings (Russell, Red Brick, New, Old, ex-poly, etc.) which some take quite seriously. 
2. It may be an accident of the universities I looked at, but US universities seem to follow a different trend. The two at the top (MIT and Stanford) had huge per student income. It may be that the other activities they are involved in distract them from scholarship. It could also be that there is a limit to the scaling power of a university as an organisation. 
3. The US for-profits are different from the other universities. They are doing something different and are probably more similar (when viewed in this way) to community colleges (indicated on the chart as US FE). As universities, some of which award PhD degrees, they need to pull their socks up.

Department models and individual academics.

In an earlier incarnation of my working life, I was a member of an organisation called the heads of chemistry UK. While chatting with a colleague at a meeting, he described a model for running an academic department that could be called a 2 to 1 model.

Essentially the way this works is that the income from an academic department is split in half. One half goes to the University and the other half goes to the department. With their half, the university does what it does: supply estate, power and lights, deans, pro vice chancellors, vice chancellors, marketing, human resources, admin, and so forth. The other half pays academic staff salaries, consumables, and supplies.This model is sometimes called a contribution model, where a key performance indicator or KPI is a contribution of a particular percentage of income to the University.

This isn't far from many corporate models. I think business gets interested when a widget can be sold at roughly 2 times cost.

At the time, there was a report which essentially showed most chemistry and physics programmes in the UK were operating in deficit mostly due to research related issues ( topic for another time). If you are curious a copy of the report may be found here. That is the macro picture.

Let's look at this from the point of view of an individual staff member. Income from teaching comes from students in return for credit points (cp). In the UK, a student might take 120 credit points in a year. In a non-clinical science, those 120 credit points (cp) might cost  £10,483. A chemistry sub-discipline consisting of perhaps 3 staff might teach the equivalent of three 20 cp modules (classes for US readers) to 100 students in each. This works out to be roughly £500,000 of income or about £166,000 per academic. This is a reasonably conservative estimate as some academics will have grant income or be part of a successful RAE/REF submission.

What about salaries? For the rank and file (Lecturer/Senior Lecturer, Principal Lecturer, and Reader) in the NEW university sector, salaries are covered by a national framework. You can find a slightly outdated version here. (since then the top of the scale has risen to around £55,000).

What does this tell you? A typical teaching oriented academic is a reasonably productive worker, many will be delivering in excess of ~3x salary in "product." Yes, there are additional support requirements: estates, deans, marketing, admin, IT, but we don't do badly and many of us teach to larger cohorts of students than indicated here.

To figure your "productivity" find out the cost (sometimes quite different from advertised price to a student) per credit point (or related local measure of university "widget") multiply that by the credit points per module and then by the total number of students in the module. If team taught, divide the cash amount by the number of staff. Please use this number responsibly, it is an index and others are involved in the overall enterprise.

Inefficient maybe, but not on this scale: For profits vs. UKHE

Policy makers dream of a firm corporate hand bringing competition and efficiency to the university sector. This dream is encouraging alternative providers into UK higher education in hope they will provide equivalent or better services at competitive cost.  Support to “For-profit” institutions via public policy assumes businesses will manage profits and expenditure within a rational market. Reality does not always follow assumptions, so let’s compare some traditional and some “for-profit” institutions, think about costs, and ask questions. Let’s make use of readily available information that a student, investor, or policy-maker might consider.[i]

The traditional UK institutions considered here are the University of Cambridge, Manchester Metropolitan University, and the Open University (OU). These were compared to The University of Phoenix (UPX),[ii] Kaplan Higher Education (KHE)[iii], and Education Management Corporation (EDMC). 

Why these? Cambridge is an excellent research oriented university.  Manchester Met is a representative of the UK’s post-92 Universities and OU is the UK’s premier online learning provider. They served 264000 students (2010/11) with a net income of £1.97 billion. UPX, EDMC and KHE together served 635,000 students generating approximately £5.7 billion of revenue through a variety of “for-profit” Institutions.  To give an idea of the scale, in 2010/11 HEFCE distributed a total of £4.5 billion to UK Higher education institutions in support of teaching through its recurrent grant.

Consider the table below.

	UCAM	MMU	OU	UPX	EDMC	KHE
Students (Headcount)	20395	34970	208710	380800	150800	73800
Income (Million £)	1251	248	471	2736	1828	886
Teaching as % of income	17.0	82.3	87.4	91.5	92.8
Teaching income (Million £)	212	204	411	2503	1696
Income/student (£)	61339	7095	2254	7185	12119	12003
Teaching income/student (£)	10405	5836	1970	6574	11247
Part time  % of total students	9.4	21.2	99.9			57[i]

 What is in the numbers?

First, income at a university is not only from teaching. At an excellent University, teaching might be a small fraction of the total. In the period considered, income from teaching ranged from 17% (Cambridge) to 92.8% (EDMC).  Where did the remainder come from? It included, research grants, contracts, investments, and “other.” “Other” might include textbook sales, catering service income, and some large businesses. For example, Cambridge obtained 17% of income from teaching, about 40% from Cambridge Assessment and Cambridge University Press, and the remainder, approximately 43%, primarily from research related activities. 

Second, the “For-profits” showed fewer signs of efficiency than might be expected. KHE (73800 students) had income in excess of Manchester Met and OU combined (243680 students). Similarly, with three times the money as Manchester Met and OU, EDMC taught 60% FEWER students. The policy maker’s dream of efficiently run corporate providers profiting from inefficiency within the current UK University sector is “imaginative.”

Third, where did the money go in the for-profits? When measured against comparable UK institutions, £2.6 Billion out of £5.7 billion of income went somewhere. I think, a lot of this is vacuumed up in marketing (which they do very well).

 Before going into detail, let’s return to the Policy Maker’s dream. One view might be that a well run corporation could generate 10-20% profit while maintaining or lowering costs with no effect on academic standards. To test this dream  using the data here, we need to estimate the characteristics of a peer UK institution to the for-profit institutions in the table.

KHE’s part time students represented 57% of the total and 60% were taught online suggesting its activities are similar to an institution roughly midway between OU and Manchester Met.  Such an institution would on average have income of £4674/student. Ignore for the moment the notion that there might be £400-800/student (~10-20% ) profit to extract. KHE has an income per student nearly 3 times this. Let’s assume UPX and EDMC serve a similar proportion of non-traditional (part time) students. Suppose a simple business model in which KHE, EDMC, and UPX subcontract their teaching to a 50:50 hybrid of OU and MMU. This would leave a profit of £7329/student (61% of income) at KHE, £7445/student (61% of income) at EDMC, and £2511/student (35% of income) at UPX. Dividends to investors are somewhat lower than this might imply, especially against the notion that 10-20% could be squeezed from the 50:50 hybrid. The difference (£2.61 Billion) exceeded the combined income of Cambridge, Manchester Met, and OU. 

Is there inefficiency in UK higher education? Not on this scale.  UK Universities are not as inefficient as popular perception might believe. Equally, “for-profit” institutions are not as efficient and the amounts returned to shareholders seem wildly incommensurate with expectations set by efficiently run peer institutions. The public opportunity cost of the gap between the UK institutions and the “for-profit” providers (£2.61 Billion) could have funded all the activities of Cambridge, Manchester Metropolitan, the  OU AND the University of Warwick while keeping ~£200 million in reserve.  

Now, the question. When the new fee regime is introduced will UKHE still be as efficient? It will be difficult to assess this until the students who entered UKHE in 2011/2012 complete their studies. The trends may be visible when the 2012/2013 financial reports are out but it could be several years before a full picture is available.

As always, the data here are correct to the best of my understanding. Where mistakes have been made, please correct me.

---

[i] Information used here is from University and Corporate annual reports, UK HESA data on student numbers, and UK HEFCE data on recurring grant amounts for 2010/11. The latter is broken down into a number of categories. To obtain teaching income for the UK institutions, The teaching grant and fees items in the annual report were used. The “research” portion of the HEFCE recurring grant was subtracted from the sum of the two items. The research component of the HEFCE grant is linked to RAE performance. Exchange rate used was US$1.6 = £1 which was roughly correct for the 2010/11 period (the last for which financial data is available). New annual reports shoould be out soon.

[ii] The University of Phoenix data were obtained from the Apollo Group annual report.

[iii] Information on Kaplan Higher education is via its parent company’s (Washington Post Corporation) Annual report.

[iv] KHE reported that ~60% were taught online and 57% of its students attended part time in 2009. Part time student data was not presented for the 2010/11 year. I have taken this value as correct for 2010/11 as it appears in the annual report.

Reading UK University Balance Sheets

In preparation for a number of future posts, it is worth taking some time to have a look at a University balance sheet. Now is a particularly good time, because the financial year 2011/12 ended in July and universities will start to put their annual reports online.

Have a look here: http://www.admin.cam.ac.uk/univ/annualreport/2011/Statement.pdf where you will find the 2010/2011 academic year financial report for the University of Cambridge.

While you are at it look at these:
1) University of Warwick:  http://www2.warwick.ac.uk/services/finance/resources/accounts/accounts1011.pdf

2) Liverpool John Moores University: http://www.ljmu.ac.uk/fin/fin_docs/Financial_Statements_2011.pdf

Why look at these? Well, if you would know a University you need to know its financials and place them in context.

Why these? These are representative of three groups of universities: a very old (and very excellent): Cambridge;  A "Glass Plate" University: Warwick; and a "new" university: Liverpool John Moores (LJMU). Let's focus on Warwick and LJMU for the moment. If you look around in the financial data and also used some data from HEFCE on student numbers you might construct a table that looks something like this.

2010/11 Data	Liverpool John Moores	University of Warwick
Income (£ Millions)	176.2	419.1
HEFCE Teaching Grant (£ Millions)	64.6	42.0
Students	24680	28165
Home Students	22360	21815
Part Time Students (%)	25.4	37.4
UG (%)	82.7	63.8
Income/student (£)	7142	14880
HEFCE Teaching Grant/Home Student (£)	2890	1926

Data from HEFCE, HESA, and University Annual Reports. HEFCE teaching grant here is reported as the teaching portion of the recurrent grant (excludes items such as research, HEIF, etc). Both institutions report approximately 10% more grant than shown in the HEFCE tables. A 10% change in the bottom row of the table does not influence the overall conclusions.

 Why look at this? 

1) Take a look at that bottom line and note that the cost to government of teaching was not huge. The UK people were getting a very good deal in 2010/11. A Warwick education at an average cost to HEFCE  (read taxpayer) of around £2K/student is really not bad. In fact, I think that was EXCELLENT value for money. Whatever you think of new Universities, LJMU is also pretty good value for money. Keep in mind that Warwick and LJMU had the same per student funding regime so the higher per student cost reflects the distribution of students across the fee structure and the proportion of full time students.

2) No matter how you look at the data, there were a lot of students in the former prices groups C and D (Bands C and D). It is likely that the typical student was in these Bands.

Under the new Fee regime, Universities stand to have a substantially increased income. Consider a university with 10,000 full time students. If 5000 are former band C and 5000 are former band D students, the total increase as the new regime takes hold will be £24 million.(See previous post for Fee band Explanations).

Fine Print: The data and computations here are correct to the best of my knowledge. As in previous posts, please bring any errors to my attention.