Minority report

The Times Higher ran a piece today with the news that ‘Academics [were] in the minority at more than two-thirds of UK universities’. This accompanied by comments about ‘an army of administrators’ and the like.

It’s hard to know what to make of stories like this. One the one hand, they belong to the long tradition of ‘academics versus administrators’ – the twin nonsenses of universities would work better if it was just left to academics and academics don’t understand the things that they should do to make the place run better.

But there is an important point which is about the growth of universities as complex organisations, with increasing regulation and accountabilities, and ever-more-demanding students.  Is it surprising that as universities get bigger and have to do more complex things, many of them decide that it’s best to hire staff with the right specific skills and abilities, allowing academic staff to focus on teaching and research? Napoleon said – in French, I expect – that armies march on their stomachs. Perhaps in the modern university academics teach with their professional service colleagues.

The THES data is easily replicable – it’s the staff data from Table 5 of HESA 2013-14, and the Finance data from Table 7.  The story identifies the top and bottom ten universities for costs per non-academic staff member, and percentage. It’s worth looking at the missing data – the top and bottom ten universities for costs per academic staff member. Here they are, using the same THES definition of only counting universities with more than 500 academic staff:

Highest ten costs per academic

Institution	Average cost - academic	Average cost - support
London Business School	£252,713	£52,334
London School of Economics	£85,298	£46,051
City University, London	£76,650	£51,701
UCL	£74,547	£42,560
King’s College London	£72,078	£51,149
Imperial	£68,464	£49,937
Royal Holloway	£67,837	£38,038
University of Reading	£65,305	£34,201
Brunel University, London	£64,663	£45,046
Strathclyde University	£63,891	£34,076

Lowest ten costs per academic

Institution	Average cost - academic	Average cost - support
University of the Arts, London	£32,167	£63,656
University of Kent	£39,152	£35,133
Coventry University	£41,234	£44,136
Edge Hill University	£41,503	£32,735
University of Sunderland	£42,020	£36,023
Southampton Solent University	£42,060	£49,961
University of Chester	£42,105	£31,879
University of Central Lancashire	£44,669	£53,123
Anglia Ruskin University	£45,600	£50,682
Staffordshire University	£46,010	£41,408

I’d observe two things from these lists: firstly, higher academic staff costs seem to have an association with London location – Strathclyde being an outlier here. And lower unit costs seem to be associated with newer universities.

And then a further observation – the average cost per support staff does not seem to follow this pattern, with high and low in both of my lists, and a mix of locations and types amongst the THES lists.

So perhaps this says that there isn’t yet a standard business model for university professional support services. Now that’s an issue worth thinking about – if there is a best way to do it, the benefits for universities and their staff are large – efficiency and effectiveness, yes, but also greater transferability of skills, and possibly better career progression for professional service staff.

Staff costs

I’m looking at the issues around managing costs in higher education at the moment (look out for a post in the next few days about why not all vacant posts can be replaced). My starting point has been to look at the data, and in particular the high level data on staff costs.

HJ calculation from HESA data

The chart – which I have calculated using HESA data – shows the proportion of universities’ total income which is spent on staff.  (See my post from October 2014 which explains what this means and why it matters). There is data for the four home nations plus UK-wide data, covering the period from 1994-95 to 2013-14, being the most recent HESA data set available.

The chart tells a story of good financial times in the early years of the Labour government in the late 90’s; and closer management of spend following the introduction of top-up fees in 2006.

Of the four national patterns, England accounts for by far the largest share of the overall income, so it’s no surprise that the total UK and the England lines follow each other closely. In turn, Scotland appears over time to track England more closely. The data for Northern Ireland looks peaky, but don’t forget that it’s a very small sector (only four institutions, of which two are very small), so individual institutional strategy will have a disproportionate impact. Wales looks to be going against the grain – increasing the proportion of income spent on staff compared to the other nations.

With devolved funding this can happen – this may reflect differences in funding for capital, for instance, rather than deliberate policy by Welsh universities to grow staff spend. But HEFCW in Llanishen might be interested to find out why, before the Assembly Finance Committee in Cardiff Bay asks the same question.

Key Performance Indicators

Look on the webpages of almost every university, tucked away sometimes in the ‘About Us’ pages, and you’ll find a section on KPIs – Key Performance Indicators. I thought it might be useful to write a bit about what they are, and why they matter. In this post I’ll also go into a bit of detail about one common KPI – staff costs as a percentage of income.

A performance indicator simply means a measure of how you’re doing. So, if I’m typing, words per minute might be a performance measure. But so might typos per sentence. Or proportion of screen-breaks properly taken. Performance indicators are quite easy to come up with (although, as a game, making up performance indicators does get pretty tedious after a while).

Now think about all the things that go on in a university (or any other organisation, for that matter), and that can be measured, and you’ll see that a full list of university performance indicators would be huge. And so Key Performance Indicators are those chosen for particular monitoring.

Chosen is an important point here. There’s no absolute list of what matters to an organisation: it is a judgment about strategy and circumstances. There are KPIs which are measured by university regulators (funding councils, research councils etc), and they do tend to get measured by universities, but it is a choice that universities make. Sometimes a pretty obvious choice, but a choice nonetheless.

The choice is made by governing bodies, on the advice of the VC and colleagues in the senior management team. And the choice tells you about what’s important to the University. If a university is trying to improve its student experience, it might have staff:student ratio as a KPI. Or proportion of space of the highest standard. Or NSS results. But if it has a financial focus in its strategy (for investment, or to ensure sustainability) it might choose income per student (or cost per student).

The other jargon term is metric. That’s the actual value associated with a Key Performance Indicator. So for my words-per-minute typing KPI, the metric at the moment is about twenty. And the typos-per-sentence is about two. Could do better! And that’s the point – there’s often a target (or a range) associated with a KPI and metric. I’d like to type with fewer errors and more words per minute. So using a KPI I can measure my progress.

So that’s the theory. Let’s take a common university KPI – proportion of income spent on staff – and look at that in more detail.

The definition is straightforward. Take a university’s total income and its total spend on staff, both of which are in the annual financial statement, and divide the latter by the former. So if my university’s income is £100m per year and staff costs are £55m per year, then the proportion of income spent on staff KPI has a value of 55/100, or 55%.

Here’s two charts which show the KPI. They’re both drawn from 2012-13 HESA data. The first is a scatter plot, where each university is represented by a single dot.

This shows that for most universities, staff costs are in the range of forty-something to sixty-something percent of income, with an average around fifty-something percent. It also shows there are some outliers – some very large institutions, with turnovers of over one billion pounds (and I do hope that the Vice-Chancellors report that to their governing bodies in the style of Dr Evil); and some smaller institutions with much smaller (3% and 14%) spend on staff. These latter are instructive: one is a small consortium institution, and doesn’t directly employ many staff at all; the other is a university with a huge geographical spread and lots of costs associated with teaching over this area. It shows why the choice of KPIs depends on the specific circumstances of the institution. This probably isn’t a very useful KPI for those particular institutions

Now let’s see a different chart, using the same data. This is a bar chart where the institutions are arranged from left to right in ascending order of turnover. That is, the bar furthest to the left is the institution with the smallest turnover, the bar furthest to the right is the institution with the largest turnover. The height of each bar shows the proportion of income spent on staff.

This is more like a range of mountains – a few really high peaks and a few deep valleys, but most are around the 45%-55% range. What’s interesting here, I think, is that there’s no obvious economy of scale – larger institutions don’t seem to have a significantly lower proportionate spend on staff. This is counterintuitive.

So why does it matter? Staff costs are unlike other costs for a university. Firstly, they are long term. Once you’ve employed someone then you need to keep paying them, month after month, and to stop paying them involves effort and more expense. Secondly, they increase on a regular basis. Pay increases incrementally for many, simply by the passage of time as you go up the pay scale; there are pay increases every year through negotiation with Trades Unions; and on-costs – pension contributions and national insurance contributions – go up not down, particularly at the moment. A pound spend on employing someone this year becomes more like £1.05 next year, and so on. To employ someone is to make a commitment. 

And if you’re managing an institution, you need to know about this. If you have to find money each year for a rising staff bill, then you have fewer opportunities for other things – equipment, buildings, training costs, IT, library books, student bursaries and scholarships. So an institution which can manage its staff costs has more opportunities. And when there’s a large increase – a hike in pension contributions, or a big pay award (they did used to happen!) then institutions with a large staff cost as a proportion of income get hit hardest.

There can’t be an absolute target, of course. Some subjects cost more to teach than others – for instance clinical subjects. Some subjects need a lot of equipment, which keeps the proportion of spend on staff low because lots of money is spent on kit. And institutions make choices about how they operate which make an impact – for instance, having small tutorial groups.

There’s no moral to this tale: if you have a high proportion of staff spend, and know why, and it fits with your strategy, maybe that’s fine. And too low a proportionate spend and maybe your institution isn’t doing things as well as it might – universities are people businesses. But the one number does help you understand quite a lot about an institution’s plans and how sustainable it is (that is, how resilient to shocks, or how much scope for investment it has). And if you’re accountable for the institution, or deciding who to give more resource to, those are good things to know.