Size can be bad for your (college financial) health?

The KPMG report on Delivering Value for Money through Infrastructural Change found “evidence that larger colleges may be more efficient and have the advantage of economies of scale”. The graphs certainly show administrative costs fall with overall costs. This finding was a key underlying hypothesis in the report. More perplexingly the report also found that when comparing the percentage of General FE (GFE) colleges’ surpluses against their total income: “It shows a slight negative correlation between total income and surplus as a proportion of this. This is an emerging finding which we will be exploring further.”

The report does not really explain this peculiar finding that larger GFE colleges have worse financial performance despite lower admin costs. The demise of the LSC and the squeeze on consultancy fees may have put the kibosh on a further exploration.

Was the analysis of the 2007/8 college accounts a historical anomaly? Maybe not. I cranked the numbers for GFE colleges using the 2008/9 accounts. (As I suspect that Greater London with its high costs and extra funding may distort the picture, I excluded London colleges. I also omitted Newcastle College which is so much larger than the rest of the sector.) What I found was a slight deterioration in operating surplus as a percentage of income as income increased.

How can this be that larger GFE colleges have much lower admin costs and slightly lower surpluses? Maybe these colleges are shifting resources from admin to the front line. There is some evidence for this as success rates and inspection results suggest larger colleges perform better non-financially. (Of course, larger colleges are better able to prepare for inspection which may not always mean that the outcomes are so much better.)

Another explanation might be that larger colleges suffer other diseconomies of scale which eat up much of the benefit of lower admin costs.

I do also have a nagging doubt about the data. The KPMG report shows a huge dispersion in admin costs at almost level of overall costs (their proxy for college size). Have we measurement problem? The KPMG report writers sensibly included plenty of caveats in their report – not least on the unvalidated nature of the data that they were given to analyse. In my experience of looking at benchmark data in colleges and in other sectors such as housing, accountants have difficulty in categorising costs even when there is clear guidance.

Do problems in dividing costs between admin and other categories explain why we have these much lower admin costs and slightly lower surpluses? I wonder if these problems do explain some of the peculiar findings – maybe larger colleges, as a consequence of size, have difficulty in identifying the admin costs in the remote reaches of their empires and in distinguishing admin from other support in curriculum areas.

Any other explanations?